Integrated Resource Plan - TVA

Integrated Resource Plan 2 015 F I N A L R E P O R T

T E N N E S S E E VA L L E Y AU T H O R I T Y

Message from the CEO

TVA is pleased to publish the 2015 Integrated Resource Plan (IRP) that provides direction for how TVA will meet the long-term energy needs of the Tennessee Valley region. This document and the associated Supplemental Environmental Impact Statement evaluate scenarios that could unfold over the next 20 years. It discusses ways that TVA can meet future electricity demand economically while supporting TVA’s equally important mandates for environmental stewardship and economic development across the Valley. And it reinforces that TVA is headed in the right direction. As in the 2011 version, this report indicates that a diverse portfolio is still the best way to deliver low-cost, reliable electricity to those we serve. The 2015 IRP was developed in collaboration with TVA customers and consumers, business and industry leaders, community, academia, environmental advocates, and everyday people who depend on the public power TVA provides in partnership with local power companies. The involvement of so many passionate and informed people ensured thorough consideration of the differing views, priorities and issues that co-exist in the Valley. Special thanks go to the IRP Working Group and the Regional Energy Resource Council for their efforts to help create a more robust strategy that incorporates all resources to support load growth and serve our customers. The 2015 IRP is an important document that will assist TVA in fulfilling its mission of serving the people of the Valley to make life better by delivering safe, clean, reliable and affordable electricity; a cleaner environment and sustainable economic opportunities. We thank our many partners for their contribution to this effort.

William D. Johnson

I N T E G R AT E D R E S O U R C E P L A N – 2 0 1 5 F I N A L R E P O R T

Table of Contents

Executive Summary�� 1 Chapter 1 - TVA’s Energy Future�� 7 1.1 TVA Overview�� 8 1.1.1 TVA’s Mission�� 8 1.1.2 TVA’s Customers�� 9 1.2 Integrated Resource Planning�� 9 1.2.1 IRP Objectives��10 1.2.2 IRP Development��10 1.3 Supplemental Environmental Impact Statement��10 Chapter 2 - IRP Process��11 2.1 Scoping��12 2.2 Develop Study Inputs and Framework ��12 2.3 Analyze and Evaluate ��14 2.4 Present Initial Results and Gather Feedback��14 2.5 Incorporate Feedback and Perform Additional Modeling��14 2.6 Identify Target Power Supply Mix��14 2.7 Approval of IRP Recommendations��14 Chapter 3 - Public Participation��15 3.1 Public Scoping Period��16 3.1.1 Public Meetings and Webinars��17 3.1.2 Written Comments��17 3.1.3 Results of the Scoping Process��17 3.1.4 Additional Comments��18 3.2 Public Involvement in Developing Study Inputs and Framework ��18 3.2.1 IRP Working Group Meetings��18 3.2.2 Public Briefings��19 3.3 Public Involvement in Review of the Draft IRP��19 3.3.1 Public Meetings �� 20 Chapter 4 - Need for Power Analysis�� 25 4.1 Estimate Demand�� 26 4.1.1 Load Forecasting Methodology�� 26 4.1.2 Forecast Accuracy�� 28 4.1.3 Forecasts of Peak Load and Energy Requirements�� 28 4.2 Determine Reserve Capacity Needs�� 29 4.3 Estimate Supply�� 30 4.3.1.1 Baseload Resources��30 4.3.1.2 Intermediate Resources�� 31 4.3.1.3 Peaking Resources�� 31 4.3.1.4 Storage Resources�� 31

4.3.2 Capacity and Energy��31 4.3.3 Current TVA Capacity and Energy Supply�� 32 4.4 Calculate the Capacity Gap�� 33

i


Table of Contents

Chapter 5 - Energy Resource Options�� 35 5.1 Energy Resource Selection Criteria�� 36 5.1.1 Criteria for Considering Resource Options�� 36 5.1.2 Characteristics Required for Resource Options�� 36 5.2 Resource Options Included in IRP Evaluation�� 37 5.2.1 Existing Assets by Resource Category�� 37 5.2.2 New Assets by Resource Category�� 40 Chapter 6 - Resource Plan Development and Analysis�� 51 6.1 Development of Scenarios and Strategies �� 52 6.1.1 Development of Scenarios�� 52 6.1.2 Development of Planning Strategies �� 59 6.2 Resource Portfolio Optimization Modeling�� 62 6.2.1 Development of Optimized Capacity Expansion Plan �� 63 6.2.2 Evaluation of Detailed Financial Analysis �� 63 6.2.3 Uncertainty (Risk) Analysis �� 64 6.3 Portfolio Analysis and Scorecard Development �� 66 6.3.1 Selection of Metric Categories�� 66 6.3.2 Development of Scoring and Reporting Metrics�� 67 6.3.3 Scorecard design�� 72 6.4 Strategy Assessment Process�� 73 Chapter 7 - Study Results�� 75 7.1 Analysis Results��76 7.1.1 Firm Requirements and Capacity Gap��76 7.1.2 Expansion Plans �� 77 7.1.3 The Clean Power Plan and the IRP�� 89 7.2 Scorecard Results�� 90 7.3 Scoring Metric Comparisons�� 93 7.4 Preliminary Observations�� 94 Chapter 8 - Strategy Assessments�� 95 8.1 Strategy Assessments�� 96 8.1.1 Cost and Risk Assessment�� 96 8.1.2 Environmental Stewardship��101 8.1.3 Flexibility��102 8.1.4 Valley Economics��103 8.1.5 Summary of Initial Observations��104 8.2 Reporting Metrics Comparisons��106 8.3 Sensitivity Analysis��107 Chapter 9 - Recommendations�� 111 9.1 Study Objectives�� 112 9.2 Findings�� 112 9.3 Developing the Recommendation�� 114 9.4 Target Power Supply Mix�� 115

ii


Table of Contents

Chapter 10 - Implementation Challenges and Next Steps��119 10.1 Overview of Next Steps��120 10.2 Implementation Challenges��120 10.3 Policy Considerations��121 10.4 Process and Modeling Improvements ��121 10.5 Conclusion��122 Appendix A - Generating Resource Cost and Performance Estimates��123 Appendix B - Assumptions for Renewables��129 Appendix C - Distributed Generation Evaluation Methodology��135 Appendix D - 2015 IRP: Modeling Energy Efficiency��141 Appendix E - Capacity Plan Summary Charts��159 Appendix F - Method for Computing Environmental Metrics��171 Appendix G - Method for Computing Valley Economic Impacts��185

iii


Table of Contents

List of Figures Figure 1: 2014 TVA Portfolio........................................................................................................................ 2 Figure 2: Range of MW Additions by 2023 & 2033...................................................................................... 3 Figure 2-1: List of Scenarios and Strategies...............................................................................................13 Figure 3‑1: Distribution of Scoping Comments by State.............................................................................17 Figure 3‑2: Inputs and Framework Public Briefings....................................................................................19 Figure 3‑3: Draft IRP Public Meetings....................................................................................................... 20 Figure 4‑1: Comparison of Actual and Forecasted Annual Peak Demand.................................................. 28 Figure 4‑2: Comparison of Actual and Forecasted Net System Requirements........................................... 28 Figure 4‑3: Peak Demand Forecast ........................................................................................................ 29 Figure 4‑4: Energy Forecast...................................................................................................................... 29 Figure 4‑5: Illustration of Baseload, Intermediate and Peaking Resources................................................. 30 Figure 4‑6: Baseline Capacity, Summer Net Dependable MW.................................................................. 33 Figure 4‑7: Estimating the Capacity Gap................................................................................................... 33 Figure 4‑8: Capacity Gap Range.............................................................................................................. 33 Figure 4‑9: Energy Gap............................................................................................................................ 34 Figure 5‑1: Coal Fleet Map........................................................................................................................ 38 Figure 5‑2: Coal Fleet Portfolio Plans........................................................................................................ 39 Figure 5‑3: List of New Assets...................................................................................................................41 Figure 5‑4: Nuclear Expansion Options..................................................................................................... 42 Figure 5‑5: Coal Expansion Options......................................................................................................... 43 Figure 5‑6: Gas Expansion Options.......................................................................................................... 44 Figure 5‑7: Hydro Expansion Options........................................................................................................ 45 Figure 5‑8: Utility-Scale Storage Options.................................................................................................. 46 Figure 5‑9: Wind Expansion Options.........................................................................................................47 Figure 5‑10: Solar Expansion Options....................................................................................................... 48 Figure 5‑11: Biomass Expansion Options.................................................................................................. 49 Figure 5‑12: DR Expansion Options.......................................................................................................... 49 Figure 5‑13: EE Expansion Options........................................................................................................... 50 Figure 6‑1: Key Uncertainties.................................................................................................................... 53 Figure 6‑2: Scenario Key Characteristics.................................................................................................. 54 Figure 6‑3: Energy Demand Assumptions................................................................................................. 55

iv


Table of Contents

Figure 6‑4: Gas Price Assumptions........................................................................................................... 56 Figure 6‑5: Coal Price Assumptions.......................................................................................................... 57 Figure 6‑6: CO2 Price Assumptions.......................................................................................................... 58 Figure 6‑7: Key Planning Strategy Attributes............................................................................................. 60 Figure 6‑8: Planning Strategies Key Characteristics.................................................................................. 61 Figure 6‑9: Strategy Descriptions.............................................................................................................. 62 Figure 6‑10: Sample Stochastic Result..................................................................................................... 64 Figure 6‑11: Example Uncertainty Ranges................................................................................................. 65 Figure 6‑12: Strategic Imperatives............................................................................................................. 66 Figure 6‑13: Scoring Metrics..................................................................................................................... 68 Figure 6‑14: Scoring Metric Formulas........................................................................................................ 69 Figure 6‑15: Reporting Metrics.................................................................................................................. 70 Figure 6‑16: Reporting Metric Formulas.....................................................................................................71 Figure 6‑17: Scorecard Alignment............................................................................................................. 72 Figure 6‑18: Scorecard Template.............................................................................................................. 72 Figure 7‑1: Firm Requirements by Scenario................................................................................................76 Figure 7‑2: Range of Capacity Gaps by Scenario...................................................................................... 77 Figure 7-3: Incremental Capacity Additions for All 25 Cases...................................................................... 78 Figure 7‑4: Capacity (Summer Net Dependable Megawatts) for Strategy A by Scenario............................ 80 Figure 7‑5: Energy (Terawatt Hours) for Strategy A by Scenario................................................................ 80 Figure 7‑6: Capacity (Summer Net Dependable Megawatts) for Strategy B by Scenario........................... 82 Figure 7‑7: Energy (Terawatt Hours) for Strategy B by Scenario................................................................. 82 Figure 7‑8: Capacity (Summer Net Dependable Megawatts) for Strategy C by Scenario........................... 84 Figure 7‑9: Energy (Terawatt Hours) for Strategy C by Scenario................................................................ 84 Figure 7‑10: Comparison of Energy Efficiency Resources in Strategy D..................................................... 85 Figure 7‑11: Capacity (Summer Net Dependable Megawatts) for Strategy D by Scenario.......................... 86 Figure 7‑12: Energy (Terawatt Hours) for Strategy D by Scenario............................................................... 86 Figure 7‑13: Comparison of Renewable Resources in Strategy E.............................................................. 87 Figure 7‑14: Capacity (Summer Net Dependable Megawatts) for Strategy E by Scenario.......................... 88 Figure 7‑15: Energy (Terawatt Hours) for Strategy E by Scenario .............................................................. 88 Figure 7‑16: Strategy A Scorecard............................................................................................................ 90 Figure 7‑17: Strategy B Scorecard............................................................................................................. 91 Figure 7‑18: Strategy C Scorecard............................................................................................................ 91

v


Table of Contents

Figure 7‑19: Strategy D Scorecard............................................................................................................ 92 Figure 7‑20: Strategy E Scorecard............................................................................................................ 92 Figure 7‑21: Scoring Metrics by Strategy & Scenario................................................................................. 94 Figure 8‑1: System Average Cost.............................................................................................................. 96 Figure 8‑2: Total Plan Cost (PVRR)............................................................................................................ 97 Figure 8‑3: Risk/Benefit Ratio................................................................................................................... 98 Figure 8‑4: Risk Exposure........................................................................................................................ 99 Figure 8‑5: Cost/Risk Trade-Offs............................................................................................................ 100 Figure 8‑6: Environmental Impacts..........................................................................................................101 Figure 8‑7: System Regulating Capability.................................................................................................102 Figure 8‑8: Valley Economics...................................................................................................................103 Figure 8‑9: Reporting Metrics by Strategy & Scenario..............................................................................107 Figure 8-10: List of Sensitivity Cases........................................................................................................108 Figure 9-1: 2014 TVA Portfolio.................................................................................................................. 113 Figure 9-2: Evaluated MW Additions/Retirements by 2033 from IRP Base and Sensitivity Case Analysis......................................................................... 114 Figure 9-3: Range of MW Additions by 2023 & 2033............................................................................... 116 Figure 10.1: Remaining IRP Process Steps...............................................................................................120 Figure B‑1: Example of Wind Monthly-mean variability of net power capacity by 3TIER............................131 Figure B‑2: Sites across Tennessee Valley with historical solar irradiance data supplied by CPR .............132 Figure B‑3: Solar Fixed Axis and Utility Tracking Capacity Factors by Month............................................132 Figure B‑4: NDC by hour of the top 20 peak load days of Summer 1998-2013........................................132 Figure C‑1: DG Market Segments and Penetration Levels across IRP Scenarios......................................136 Figure C‑2: Correlation of IRP CO2 uncertainty values to EIA source data...............................................137 Figure C‑3: Development of National Renewable DG Penetration Levels..................................................138 Figure C‑4: National Renewable Energy Adoption Levels (Utility-led and DG)...........................................138 Figure C‑5: Residential/Commercial DG Adoption Levels (by 2040).........................................................139 Figure C‑6: Residential/Commercial DG Adoption Levels (Annual)...........................................................139 Figure D‑1: Energy Efficiency Performance on a Typical Peak Summer Day (2023)...................................143 Figure D‑2: Energy Efficiency Monthly Profile (2023).................................................................................144 Table D‑4: Tier Step Changes..................................................................................................................147 Figure D‑3: Levelized EE block cost comparison ($/MWh) compared to a greenfield combined cycle plant over time..........................................................................147

vi


Table of Contents

Figure D‑4: Example of a residential audit program modeled as part of a residential block......................152 Figure D‑5: Planning Factor Adjustment over time....................................................................................154 Figure D‑6: Uncertainty bands in $/MWh for each of the EE sector blocks as compared to a greenfield combined cycle plant................................................................156 Figure D‑7: Levelized Cost comparisons in 2015 (2015$/MWh)................................................................157 Figure D‑8: Levelized cost comparison ($/MWh) of EE tiers in 2015 and 2033........................................157 Figure D‑9: Levelized Cost Comparison by Sector through time..............................................................157 Figure F‑1: Scoring Metrics......................................................................................................................172 Figure F‑2: Scoring Metric Formulas........................................................................................................172 Figure F‑3: Reporting Metrics..................................................................................................................173 Figure F‑4: Reporting Metric Formulas.....................................................................................................173 Figure G‑1: Input and Output Impacts......................................................................................................186 Figure G‑2: Results..................................................................................................................................188 Figure G‑3: Current Outlook....................................................................................................................189 Figure G‑4: Stagnant Economy................................................................................................................191 Figure G‑5: Growth Economy..................................................................................................................193 Figure G‑6: De-Carbonized Future...........................................................................................................195 Figure G‑7: Distributed Marketplace.........................................................................................................197 Figure G‑8: Nonfarm Employment...........................................................................................................199

List of Tables Table 8‑1: Summary of Observations by Metric Category��104 Table 8‑2: Summary of Observations by Strategy ��105 Table D‑1: Tier, Sector, and Block Hierarchy��145 Table D‑2: Weighting of EE Programs ��145 Table D‑3: Net to Gross ratios and Lifespans for the EE programs within sectors��146 Table D‑5: Block Characteristics for each sector��148 Table D‑6: Resource Characteristic Comparison with EE��150 Table D‑7: Design and Delivery Uncertainties��151 Table D‑8: Indirect and Direct Stochastic Variables��155

vii


Table of Contents

Acronym Index IRP

Integrated Resource Plan

TVA

Tennessee Valley Authority

EE

Energy Efficiency

IRPWG

Integrated Resource Plan Working Group

PPA

Purchase Power Agreement

SMR

Small Modular Nuclear Reactor

LPC

Local Power Company

EIS

Environmental Impact Statement

SEIS

Supplemental Environmental Impact Statement

NEPA

National Environmental Policy Act

EEDR

Energy Efficiency and Demand Response

FY

Fiscal Year

MW Megawatt MWh

Megawatt hour

GWh

Gigawatt hour

TWh

Terawatt hour

kWh

Kilowatt Hour

MAPE

Mean Absolute Percent Error

CC

Combined Cycle gas plant

CT

Combustion Turbine

CO2

Carbon Dioxide

EPU

Extended Power Uprates

PWR

Pressurized Water Reactor

APWR

Advanced Pressurized Water Reactor

HVDC

High Voltage Direct Current

IGCC

Integrated Gas Combined Cycle

SCPC

Supercritical Pulverized Coal

CCS

Carbon Capture and Sequestration

PVRR

Present Value of Revenue Requirements

viii

Executive Summary Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

1

Re-evaluate

Recommend


Executive Summary

Overview

Developing the Recommendation

The Tennessee Valley Authority’s 2015 Integrated Resource Plan (IRP) will guide TVA in making decisions about the energy resources used to meet future demand for electricity through 2033. More like a compass than a GPS, it provides broad direction but not street-by-street instructions.

The recommendation takes into account customer priorities around power cost and reliability across different futures. Implementing the least-cost resource plan with these priorities in mind will help ensure TVA continues to fulfill its mission to serve the people of the Tennessee Valley.

The recommendation in this IRP provides strategic guidance on the resource mix to successfully respond to changing market conditions. The resource mix will be: low cost, reliable, risk-informed, diverse, environmentally responsible and flexible.

In developing a recommendation from the study, TVA has elected to establish guideline ranges for key resource types (owned or contracted) that make up the target power supply mix. This general planning direction is expressed over the twenty year study period while also including more specific direction over the first ten year period. In order to distill the considerable number of cases evaluated through the original scenario and strategy analysis and the sensitivity cases, the recommendation uses ranges that are centered on results obtained under the Current Outlook scenario. See Section 2.2, Develop Study Inputs and Framework, for a description and definition of scenarios. The other scenario results provide a sense of how the recommended mix might change as the future changes. The need to shift the resource mix will be based on these key variables:

This study reinforces the importance that TVA’s power be reliable, affordable and sustainable into the future. Our resource additions will build on TVA’s existing diverse asset portfolio reflected in Figure 1.

Figure 1: 2014 TVA Portfolio Our results show no immediate needs for new base load plants after Watts Bar Unit 2 comes online and uprates are completed at Browns Ferry nuclear plant. Instead, we can rely on additional natural gas generation (combined cycle and combustion turbine), greater levels of cost-effective energy efficiency, and increased contributions from competitively priced renewable power. We also expect to have less coalbased generation in our energy mix than we do today, although it will continue to play an important role in the portfolio.

• Changes in the load forecast • The price of natural gas and other commodities • The pricing and performance of energy efficiency and renewable resources • Impacts from regulatory policy or breakthrough technologies 2


Executive Summary

Target Power Supply Mix

The first three variables represent the fundamental drivers for most of the variation in the resource plans produced across the strategy/scenario combinations. Our planning direction, while initially focused around the current view of the future, is flexible enough to indicate how that power supply mix would shift if one or more of these key variables exhibits a material change from the forecasts used in the IRP. We also recognize that impacts from breakthrough technologies (like a significant advance in energy storage) would be a game-changer, and TVA will continue to monitor emerging technology as it develops.

The recommendation for the power supply mix is presented in the form of ranges around boundaries established by the IRP results. The solid bars represent the recommended range from the IRP scenario that best represents our current estimation of the future. However, recognizing that a variety of future scenarios are possible, the wider range (shown in horizontal black lines) is provided to represent how the resource portfolio may respond in different future scenarios. The recommended ranges represent incremental additions (or retirements) from the existing resource fleet and may be owned or contracted assets. The results are bounded by the full range of the IRP cases and sensitivity runs which reaffirm the merits of a diverse portfolio. TVA will closely monitor key input variables including changes in the load forecast, the price of natural gas and other commodities, the pricing and performance of energy efficiency and renewable resources, and impacts from regulatory policy or breakthrough technologies to help determine whether adjustments should be made to the recommended ranges.

This approach provides a more robust recommendation than was developed in the 2011 IRP. While that approach provided a solid framework for the resource decisions TVA has made since the TVA Board accepted the IRP planning direction in the spring of 2011, the changing utility marketplace requires a more flexible guide that provides decision-makers with a clear understanding of how the resource mix would evolve in response to future uncertainties. The recommendation meets the dual objective of ensuring flexibility to respond to the future while providing guidance on how our resource portfolio should change as the future unfolds.

Figure 2: Range of MW Additions by 2023 & 20331 1 MWs are incremental additions from 2014 forward to align to the IRP analysis base year. Board-approved coal retirements and natural gas additions as of August 2015 are excluded from the totals.

3


Executive Summary

Figure 2 shows the boundaries of resource additions we are proposing by the end of the first 10 years of the study (2023) and by the end year of the study (2033), shown in Megawatts (MW). The specific recommendations by resource type are summarized below.

Wind: Add between 500 and 1,750 MW by 2033, depending on pricing, performance, and integration costs. Given the variability of wind selections in the scenarios, evaluate accelerating wind deliveries into the first 10 years of the plan if operational characteristics and pricing result in lower-cost options.

Coal: Continue with announced plans to retire units at Allen, Colbert, Johnsonville, Paradise and Widows Creek. Evaluate the potential retirement of Shawnee Fossil Plant in the mid-2020s if additional environmental controls are required. Consider retirements of fully controlled units if cost-effective.

Natural Gas (Combustion Turbine and Combined Cycle): add between 700 and 2,300 MW by 2023, and between 3,900 and 5,500 MW by 2033. The key determinants of future natural gas needs are trajectories on natural gas pricing and energy efficiency and renewables pricing/ availability.

Nuclear: Complete Watts Bar Unit 2 and pursue additional power uprates at all three Browns Ferry units by 2023. Continue work on Small Modular Reactors as part of Technology Innovation efforts and look for opportunities for cost sharing to render these more cost-effective for our ratepayers.

TVA’s recommended planning direction reaffirms its commitment to a diverse resource portfolio guided by the least-cost system planning mandate. The ranges above provide a general guideline for resource selection, but the full case analysis studied in the IRP and the SEIS includes ranges much broader than shown above driven by key drivers such as material changes in economic conditions or regulations. We believe meeting our future needs in accordance with the resource technologies and ranges in this recommendation will position TVA to continue to deliver reliable, low-cost, and cleaner power to the people of the Tennessee Valley.

Hydro: Pursue an additional 50 MW of hydro capacity at TVA facilities and consider additional hydro opportunities where feasible. Demand Response: Add between 450 and 575 MWs of demand reduction products by 2023 and similar amounts by 2033, depending on availability and cost of this customerowned resource.

How Did We Get Here? We took the findings in the initial study and the subsequent additional information gained from sensitivity analysis to form the recommendation. We used five key measures to evaluate the performance of the 25 initial plans created as part of the study. Those results told us:

Energy Efficiency: Achieve savings between 900 and 1,300 MW by 2023, and between 2,000 and 2,800 MWs by 2033. Work with LPCs to refine delivery methods, program designs and program efficiencies, with the goal of lowering total cost and increasing deliveries of efficiency programs.

• Cost: Total costs are similar for many of the cases over the long-term, and strategies that allow for a diverse mix of resource additions have a lower cost than those that emphasize particular technologies. Higher amounts of energy efficiency may create an upward pressure on rates in future years due to reduced sales.

Solar: Add between 150 and 800 MW of large-scale solar by 2023, and between 3,150 and 3,800 MW of largescale solar by 2033. The trajectory and timing of solar additions will be highly dependent on pricing, performance and integration costs.

4


Executive Summary

• Financial Risk: Risks are minimized by maintaining a diverse portfolio and not over-emphasizing any specific resource type.

and performance assumptions are critical to the final result, and lower costs or less uncertainty around this resource would increase its selection in the portfolio. Our study results also highlight that higher volumes of Energy Efficiency tend to increase system average costs. TVA and our LPC partners will need to balance Energy Efficiency amounts and programs to ensure that those who cannot participate in these programs are not disproportionately impacted.

• Environmental Stewardship: All strategies show significant improvement in TVA’s environmental footprint and position the Tennessee Valley to have continued reductions in CO2 emissions. Strategies that emphasize energy efficiency or renewables have the best environmental results.

• Renewable selection is highly dependent on gas price, load and cost and performance assumptions.

• Valley Economics: All strategies have a similar impact on overall economic health and contribute to a strong, vibrant economy across the region.

• Natural gas pricing and load levels remain key sensitivities for all resource decisions.

• Flexibility: System flexibility is generally equivalent in most cases but is reduced when renewables are strongly emphasized.

The results of these analyses supported the ranges established in the initial findings. The sensitivity cases, coupled with the original 25 case results, provide a robust set of 2,000 potential resource additions evaluated in the IRP from which the final recommendation was derived.

Reviewing these results led to questions from stakeholders about how changes in assumptions or resource choices might impact the findings. A series of sensitivity cases were evaluated with five main categories: nuclear additions, modified assumptions for energy efficiency, alternative renewable resource costs, impacts associated with forcing resources not otherwise selected into the mix, and changes in fundamental drivers such as load growth and fuel pricing.

Policy Considerations In the process of developing the cases and reviewing the results with stakeholders, a number of policy-related issues were raised that are outside the scope of the IRP itself but will need to be considered as we move toward implementation of any recommendations from the study.

Sensitivity Cases Key Findings

For example, we recognize that a commitment to significant levels of energy efficiency as part of the resource portfolio will likely put upward pressure on rates (absent a redesign) that could have negative consequences for low/fixed income customers as well as renters. The details of the approach we might take are outside the scope of the IRP report, but the study work we have completed will inform the followon planning and evaluation of the Energy Efficiency portfolio. We also know that program design will be a key challenge to ensure that the broadest possible Energy Efficiency portfolio can be offered through the LPCs to minimize possible bill impacts on nonparticipants.

• New nuclear or coal assets would offset gas builds and renewable purchases. Nuclear additions increase total cost but lower fuel risk. Small Modular Reactors are presently cost-prohibitive, but cost-sharing would render them more financially attractive. Subsequent IRPs will need to address the expected expiration of licenses for TVA’s operating nuclear units which may occur beyond the present study window. • Energy Efficiency was successfully modeled as a selectable, supply-side equivalent resource. In general, energy efficiency programs eliminate the need for natural gas units as well as some renewable purchases. As with any resource, cost

5


Executive Summary

We also realize that electric rates and job growth are critical concerns for Valley residents. While we have chosen to focus on two specific metrics to assess the macro-economic impacts of our resource choices, TVA remains committed to our least-cost mandate and our responsibility for regional economic development. Although the IRP itself does not analyze either of these issues, the findings in this planning study do become key inputs in the financial planning cycle that helps TVA set rates and fund economic development activities. There are several other policy issues that come into play when implementing recommendations from the IRP, especially if the target power supply mix relies on more load-side options, like Energy Efficiency programs, or resources that are more dispersed, like wind or solar facilities. Because of our unique business model, TVA and Local Power Companies (LPCs) will have to collaborate in new and innovative ways to ensure that this evolving resource portfolio remains reliable and provides the most value to all customers.

6

Chapter 1

TVA’s Energy Future Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

7

Re-evaluate

Recommend


Chapter 1: TVA’s Energy Future

1 TVA’s Energy Future

1.1 TVA Overview

The 2015 Integrated Resource Plan (IRP) will guide TVA in making decisions about the energy resources we will use to meet future demand for electricity in the Tennessee Valley. Having a long-range energy resource plan enables us to provide affordable, reliable electricity to the people we serve. It is crucial in a constantly changing business and regulatory environment and will better equip us to meet many of the challenges facing the electric utility industry.

1.1.1 TVA’s Mission TVA was created by Congress in 1933, and charged with a unique mission: to improve the quality of life in a seven-state region through the integrated management of the region’s resources. To help lift the Tennessee Valley out of the Great Depression, TVA built dams for flood control, provided low-cost power and commercial shipping, restored depleted lands, and raised the standard of living across the region. As times have changed, we have changed with them, meeting new challenges and bringing new opportunities. Today, TVA continues to serve the people of the Tennessee Valley through its work in three areas: Energy, the Environment and Economic Development.

A key challenge is projecting how much power we will need, when and where, and identifying the optimum mix of energy resources to meet future power demand. Electricity can’t be stored economically in sufficient quantities, so electric utilities must constantly balance power supply and demand. Energy efficiency programs can help reduce that demand, and various energy resources can be used to supply future demand – from constructing new generation facilities to contracting with others to provide needed electricity, including renewable generation. But all of these options take time to implement. Given the long lead times required to plan, permit and build generating facilities, demand forecasts involve 10- to 20-year outlooks.

Energy Safe, clean, reliable and affordable electricity powers the economy of our region and enables greater prosperity and a higher quality of life for everyone. After safety, our top priority is keeping our electric rates as low as feasible and our reliability as high as possible. TVA operates the nation’s largest public power system, including 41 active coal-fired units, six nuclear units, 109 conventional hydroelectric units, four pumped-storage units, 87 simple-cycle combustion turbine units, 11 combined cycle units, five diesel generator units, one digester gas site and 16 solar energy sites.2 We also purchase a portion of our power supply from third-party operators under long-term power purchase agreements (PPAs).

In addition, we must take steps to ensure TVA has the transmission infrastructure to get electricity to where it is needed. We currently operate and maintain 16,000 miles of transmission lines across the Tennessee Valley region. As the population grows we must upgrade or expand this system. TVA must allow adequate time to properly study, engineer, site and plan environmental reviews to build additional transmission infrastructure.

TVA’s 16,000-mile-long transmission system is one of the largest in North America. For the past 14 years, the system achieved 99.999 percent power reliability. It efficiently delivered more than 161 billion kilowatt-hours of electricity to customers in FY 2014.

All of these activities entail varying levels of risk and uncertainties which we try to account for in our IRP analyses and energy resource portfolio. In earlier IRPs, we determined that a diversified energy portfolio is one of the best ways to reduce risks, and we have reconfirmed this in our 2015 IRP. It is important that we maintain a mix of energy resource options, including nuclear, natural gas, coal, energy efficiency, hydroelectric power and other renewables, to reduce the risks associated with relying too much on a specific fuel type.

TVA makes annual investments in science and technology innovation that enable us to be at the forefront of advances in the utility industry and help us meet future business and operational challenges. Core research activities directly support improving our generation and delivery assets, air and water quality 2 As of September 30, 2014

8



and clean energy integration. Currently, we are involved in research activities related to emerging technological advances in small modular nuclear reactors (SMRs), grid modernization for transmission and distribution systems, energy utilization technologies and distributed energy resources.

the number and quality of jobs in the Valley and to benefit the power system through smarter energy use.

1.1.2 TVA’s Customers Our relationships with our customers are crucial to providing affordable electricity to residents and businesses. As largely a wholesaler of electricity, TVA works in partnership with local power companies (LPCs) to deliver affordable, reliable electricity. We also deliver electricity directly to some customers, large industries and federal installations and exchange power with other interconnected utility systems.

Environmental Stewardship TVA manages natural resources of the Tennessee Valley for the benefit of the region’s people. We manage the Tennessee River system and associated public lands to reduce flood damage, maintain navigation, support power production, enhance recreation, improve water quality and protect shoreline resources.

LPCs make up most of TVA’s customer base and are the backbone of the region’s power distribution system. Accounting for roughly 87 percent of total TVA sales and 90 percent of total revenue, the LPCs are municipally-owned and consumer-owned (cooperative) utilities. TVA generates and delivers electricity to the LPCs, which deliver electricity to their residential, commercial and industrial customers. Municipal LPCs comprise the largest block of TVA customers. Many of the consumer-owned cooperative utilities were formed to bring electricity to once-sparsely populated rural, remote areas of the region.

TVA manages its power system to provide clean energy and minimize environmental impacts from its operations. Today, air quality across the region is the best it has been in more than 30 years. Since 1977, TVA has spent about $6 billion on air pollution controls and is investing approximately $1 billion in more control equipment at our Gallatin Fossil Plant in middle Tennessee. Emissions of nitrogen oxides (NOx) are 91 percent below peak 1995 levels and emissions of sulfur dioxide (SO2) are 95 percent below 1977 levels through 2013. TVA’s emissions of carbon dioxide (CO2) were reduced 32 percent between 2005 and 2013. We project approximately a 40 percent reduction in CO2 emissions by 2020 from 2005 levels. TVA is also reducing water use and waste production from its operations as it retires coal plants, increases generation from natural gas and renewable sources, and converts coal combustion residuals management to dry handling and storage.

Large industries and federal installations that buy electricity directly from TVA, such as Oak Ridge National Laboratory, account for 13 percent of total sales and 10 percent of total revenue. TVA’s electricity exchanges with interconnected utilities also can produce revenue. TVA power contracts govern customer relationships, including the pricing or rate structure under which power is sold. Our contracts with LPCs obligate TVA to generate and deliver enough electricity to meet their full electric load, including reserves, now and in the future.

Economic Development TVA’s large power system, diverse fuel mix and robust transmission system allows us to provide high reliability and competitive rates to attract industry to our region. During the past five years, TVA has helped attract or retain 240,000 jobs in our service territory and secure more than $30 billion in capital investment for the region through the Valley Investment Initiative program. This program, established in 2008, is designed to increase

1.2 Integrated Resource Planning The purpose of integrated resource planning is to meet future power demand by identifying the need for generating capacity and determining the best mix of resources to meet the need on a least-cost, systemwide basis. The integrated approach considers a broad

9



range of feasible supply-side and demand-side options and assesses them with respect to financial, economic and environmental impacts. The 2015 IRP will revise our 2011 IRP. We are updating the 2011 IRP earlier than planned because several of the assumptions used in its development changed. These include reduced demand for electricity and greater availability and lower cost of natural gas.

objectives) were considered as different combinations of strategies and predictions of future conditions were analyzed and evaluated. We conducted the IRP process in a transparent, inclusive manner that provided numerous opportunities for the public to learn about the project and participate in it. We met regularly with a wide range of stakeholders who served on the 2015 IRP Working Group. This group was composed of individuals representing state agencies, distributors of TVA power, industry groups, environmental and energy advocates, academia and research institutions, and business and economic development professionals. (More information about the IRP Working Group is provided in Chapter 3, Public Participation.) We believe this extensive outreach produced a better IRP and are grateful for the questions raised and the feedback and insights provided.

1.2.1 IRP Objectives The following objectives guide the development of this IRP: • Deliver a plan aligned to mandated least-cost planning principles • Manage risk by utilizing a diverse portfolio of supply and demand-side resources • Deliver cleaner energy and continue to reduce environmental impacts • Evaluate increased use of renewables, energy efficiency, and demand response resources • Ensure the portfolio delivers energy in a reliable manner • Develop the ability to dynamically model energy efficiency in the study • Provide flexibility to adapt to changing market conditions and identify significant sign posts • Improve credibility and trust through a collaborative and transparent process • Integrate stakeholder perspectives throughout the study

1.3 Supplemental Environmental Impact Statement As a federal agency, TVA must comply with the National Environmental Policy Act of 1970 (NEPA). This act requires all federal agencies to consider the impact of their proposed actions on the environment before making decisions. The NEPA process provides a structured way for analyzing alternative actions and for involving the public in the decision-making process. For the development of this IRP, the primary product from the NEPA process is a supplement to the 2011 environmental impact statement (EIS).

1.2.2 IRP Development

The EIS focuses on the potential impacts of the various IRP strategies more closely and in greater detail than do the environmental metrics presented in this IRP. The impacts of actions to implement the IRP, such as building and operating a new generating facility, will be the subject of action- and site-specific NEPA reviews.

TVA’s 2015 IRP was developed over a two-year period with extensive technical and economic analyses and significant participation from our customers and other stakeholders. We used an integrated, least-cost system planning process that takes into account the demand for electricity, resource diversity, reliability, costs, risks, environmental impacts and the ability to dispatch energy resources. Forecasts of inflation, commodity prices and environmental regulations were evaluated simultaneously to provide needed information. Constraints (corporate, strategic and environmental

This study was prepared in accordance with NEPA, Council of Environmental Quality regulations for implementing NEPA, and TVA’s procedures for implementing NEPA.

10

Chapter 2

IRP Process Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

11

Re-evaluate

Recommend


Chapter 2: IRP Process

2 IRP Process

results of these potential actions and potential future environments describe the portfolio in areas such as operations, financials, environmental impact, macroeconomics, and reliability.

TVA’s 2015 IRP process consists of seven distinct steps: 1. 2. 3. 4. 5.

Scoping Develop Study Inputs and Framework Analyze and Evaluate Present Initial Results and Gather Feedback Incorporate Feedback and Perform Additional Modeling 6. Identify Preferred Target Supply Mix 7. Approval of Recommended Plan

Our goal is to identify an energy resource plan that performs well under a variety of future conditions (e.g., a strong economy or a weak economy) thereby reducing the risk that a selected strategy or plan would perform well under one set of future conditions, but poorly under a different set of conditions. This increases the likelihood that TVA’s plan will provide least-cost solutions to future demands for electricity from its power system regardless of how the future plays out.

Public participation was integral to the process and is explained in more detail in Chapter 3. Steps 2 through 6 are explained in more detail in Chapter 6.

This decision making framework requires use of a scenario planning approach. Scenario planning provides an understanding of how the results of nearterm and future decisions would change under different conditions.

2.1 Scoping The public scoping period for TVA’s 2015 IRP began in October 2013. The objective in this initial step was to identify resource options, strategies and future conditions that merit evaluation in the IRP process. Public scoping comments covered a wide range of issues, including the nature of the integrated resource planning process, preferences for various types of power generation, increased energy efficiency and demand response (EEDR) and the environmental impacts of TVA’s power generation. The comments received helped to identify issues important to the public and to lay the foundation for the supplement to the 2011 Environmental Impact Statement that supports the 2015 IRP.

Future decisions that produce similar results under different conditions may mean that these decisions provide more predictable outcomes, whereas decisions that result in major differences are less predictable and therefore more “risky.” At the outset of our 2015 IRP process, we developed a set of five resource planning strategies that would be analyzed as part of the IRP. These planning strategies represent decisions that Strategies represent future TVA controls (e.g., business decisions over which asset additions, TVA has full control. idling coal plants, Scenarios represent future integration of conditions that TVA cannot more flexible control. resource options) as opposed to A portfolio is the intersection the scenarios of a strategy and a scenario described below, and represents a multiyear which represent energy resource plan detailing aspects of how TVA intends to meet the future that future power needs. TVA does not

2.2 Develop Study Inputs and Framework When developing a long-term plan for a power system, utilities typically use a least-cost decision making framework that focuses on a single view of the future. At TVA, we also use a least-cost decision making framework. The Integrated Resource Plan informs TVA on how potential resource portfolios could perform given different market and external conditions. The

12



control (e.g., more stringent regulations, fuel prices, construction costs).

Each scenario can be thought of as a model of a possible future. In one model, the economy might stagnate, fuel prices drop and electricity demand remains flat. In another, strong economic recovery could lead to increased fuel prices and to rapid recovery in electricity sales and long-term demand growth and increase the cost of building generating sources.

Different mixes of resource options (generating technologies and/or energy efficiency programs) formed the framework for distinct planning strategies that were assessed over the 20-year IRP planning horizon. The feasibility of each planning strategy was determined based on input from subject matter experts and stakeholders.

To better assess the robustness of the strategies evaluated for this IRP, we purposely structured these scenarios to present different challenges to the resource planning strategies. The scenarios differ from each other in key areas, such as projected customer demand, fuel prices and future economic and regulatory conditions.

Together we then developed a series of five scenarios representing alternative plausible futures to help us test the performance of the resource planning strategies under different conditions and, ultimately, to identify the strategy that might represent the most flexible approach to ensuring the lowest cost, most reliable power for our customers.

The five scenarios and five strategies are shown in Figure 2-1.

Figure 2-1: List of Scenarios and Strategies

13



2.3 Analyze and Evaluate

2.5 Incorporate Feedback and Perform Additional Modeling

After the resource planning strategies and scenarios were developed, the performance of each planning strategy was analyzed in detail across all of the scenarios. This phase of the IRP used industrystandard capacity expansion planning and production cost-modeling software to estimate the total cost of each combination of strategy and scenario. Other metrics, financial risks and environmental impacts, were developed from the cost-modeling results.

After the public comment period ended, all comments were reviewed and combined with other similar comments as appropriate. We have responded to all substantive comments either by revising the IRP or associated supplemental EIS or by providing specific answers in the final supplemental EIS.

2.6 Identify Target Power Supply Mix After review of the public comments received and any additional analysis, TVA staff identified a target power supply mix based on one or more of the planning strategies evaluated in the IRP. This general target reflects the mix of resources (supply and demand side) that best position the utility for success in a variety of alternative futures, while preserving the flexibility necessary to respond to uncertainty.

Unique resource plans or “portfolios” were developed, one for each combination of scenario and strategy. Each of the 25 portfolios represented a long-term, leastcost plan of different resource mixes that could be used to meet the region’s power needs. Every portfolio was ranked using metrics within a consistent, standard scorecard. Care was taken to note those portfolios that performed best overall, and those that performed well in most models of the future. The metrics were chosen based on importance to TVA’s mission and captured financial, economic and environmental impacts. Portfolios were analyzed for their robustness under stress across multiple scenarios, as opposed to total overall performance since metrics alone could signify good performance in one or two scenarios, but average or poor performance in all others.

2.7 Approval of IRP Recommendations No sooner than 30 days after the Notice of Availability, the associated EIS will be published in the Federal Register and the TVA Board of Directors will be asked to approve the recommendations included in the study, including the target power supply mix. The Board will decide whether to approve the recommendations presented in the study, to modify them or to approve an alternative. The Board’s decision will be described and explained in a Record of Decision.

2.4 Present Initial Results and Gather Feedback The draft 2015 IRP was released for public review and comment in March 2015. It presented a range of viable planning strategies for further consideration, and included scorecards and assessments using key metrics. As in the scoping period, TVA encouraged public comments on the draft IRP and associated supplemental EIS. These comments helped us identify public concerns and recommendations for the future operation of the TVA power system.

14

Chapter 3

Public Participation Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

15

Re-evaluate

Recommend


Chapter 3: Public Participation

3 Public Participation

The Tennessee Valley Renewable Information Exchange (TV-RIX) was established in September 2012, and was actually a result of the 2011 IRP ‘Next Steps’ to further analyze renewable technologies, business models and market trends. The group consisted of 17 members representing renewable interest groups, state governments, national/regional expertise and utility industry representatives. TV-RIX provided inputs on biomass, hydro, solar and wind resources for the IRP modeling process.

Understanding the varying needs and priorities of our 9 million stakeholders and striking a balance can be challenging, but is a key to the IRP process. Gaining that perspective is why we used a transparent and participatory approach in in developing this long-range plan. Obtaining diverse input and support for the IRP was one of our goals. We wanted to make sure those who wanted to participate, could do so. Our public involvement goals were to:

The Energy Efficiency Information Exchange (EEIX) was established in October 2013 to focus on the exchange of ideas on naturally occurring adoption rates of energy efficiency. The group had 13 members representing local power companies, state energy offices and non-government organizations. This group assisted in developing simple, flexible and cost-effective portfolios to be used in the IRP analysis and selection process.

• Engage numerous stakeholders with differing viewpoints throughout the process. • Incorporate public opinions into the development of the IRP, including opportunities to review and comment on various inputs, analyses and options being considered. • Encourage open and honest communication in order to provide a sound understanding of the process • Create public awareness and opportunities to receive feedback. • Form an IRP Working Group made up of people representing the broad perspectives of those who live and work in the Valley.

Public involvement was a particular focus throughout the IRP process, including steps 1 and 2, Scoping and Develop Study Inputs and Framework, and as part of step 4, Present Initial Results and Gather Feedback.

3.1 Public Scoping Period

The formation of an IRP Working Group was a cornerstone of the public input process for the 2015 IRP, just as it was for the 2011 study. Working Group members reviewed input assumptions and preliminary results and provided feedback throughout the process. They provided their individual views to TVA, as well as representing and keeping their constituencies informed regarding the IRP process.

To begin the 2015 IRP process, TVA announced the start of a 33-day public scoping period on October 21, 2013.

The 2015 Working Group consisted of 18 representatives from business and industry, state agencies, government, distributors of TVA power, academia, and energy and environmental nongovernmental organizations.

In addition, on October 31, 2013, TVA published a notice in the Federal Register of its intent to prepare a supplement to the 2011 IRP Environmental Impact Statement.

The Scoping period was publicized across the Tennessee Valley through news releases, advertisements and a notice on TVA’s website. Notices also were sent to people who participated in the development of TVA’s 2011 IRP.

At the start of the public scoping period, we explained why we were updating our IRP, its focus, how we would conduct the planning process, and how the results would guide our future energy resource decisions.

In addition to the IRP Working Group, individuals on two stakeholder groups provided TVA guidance and expertise on renewable energy resources and energy efficiency and demand response.

16



Our goals in conducting public scoping were to ensure: • Stakeholder issues and concerns were identified early and studied properly. • Reasonable alternatives were considered. • Uncertainties that could impact costs or performance of certain energy resources were identified. • Input received was properly considered and would lead to a thorough and balanced IRP. As part of scoping, we collected public input through public meetings, webinars and written comments.

3.1.1 Public Meetings and Webinars TVA held two public meetings as part of the scoping period. The first was in Knoxville, Tenn., on October 24, 2013, and the second was in Memphis, Tenn., on November 6, 2013. Both meetings were simulcast as webinars, available online.

Figure 3‑1: Distribution of Scoping Comments by State

3.1.3 Results of the Scoping Process The information collected during the public scoping period helped shape the initial framework of TVA’s 2015 IRP and was used to determine which resource options should be considered.

At each meeting, TVA staff described the process of developing the IRP and associated SEIS and then responded to questions from meeting attendees both in person and online.

The majority of the scoping comments were generated by the Sierra Club and Tennessee Environmental Council, consisting of email forms that thanked TVA for recent coal plant retirement decisions and urged TVA to prioritize the use of solar and wind energy, increase energy efficiency efforts, and work to reduce the local economic impacts of coal plant retirements. A smaller campaign, promoted by organizations affiliated with the regional coal industry, submitted form letters citing the abundance and stable cost of coal, the high capacity factor of coal plants, the employment provided by the use of coal, and coal’s contribution to low and stable rates.

Participants included the public, congressional, state and local officials, representatives from local power companies, non-governmental organizations and other special interest groups, and TVA employees. About 85 people attended the meetings in person or via webinar.

3.1.2 Written Comments TVA accepted scoping comments via mail, email, fax, comment cards and online comment forms. About 85 percent of the scoping comments were submitted either as email forms or form letters promoted by two advocacy campaigns.

Other comments, included: Energy Resource Options Most of the comments regarding potential energy resource options addressed the benefits and/or drawbacks of various energy options, including nuclear, coal, natural gas, solar and wind generation. Numerous comments encouraged increased energy efficiency

We received a total of nearly 1,100 comments. Figure 3-1 shows the distribution of these comments by state.

17



3.1.4 Additional Comments

efforts while a small number of comments encouraged increased use of other demand reduction options such as demand response and combined heat and power. Several commenters requested that TVA fully and fairly evaluate all potential energy resources

After the close of the scoping period, TVA received comments related to the IRP from two advocacy campaigns. In the spring of 2014, TVA received nearly 1,000 postcards through a Tennessee Sierra Club campaign. The message on these cards was similar to that of the Sierra Club/Tennessee Environmental Council email campaign during the public scoping period.

Environmental Impacts of Power System Operations Many of the comments addressed the negative and/or beneficial environmental and economic impacts of the use of various energy resource options. These included air pollutants, greenhouse gas emissions and climate change, spent nuclear fuel and disposal of coal ash. Several comments also mentioned impacts resulting from mining, particularly surface mining and the use of coal and hydraulic fracturing to produce natural gas. Others encouraged TVA to assess the vulnerability of its power system to climate change, to assess the effects of climate change on TVA’s power demand forecasts, and to conduct more detailed analyses of local and regional economic impacts, including employment.

In the fall of 2014, TVA received about 4,500 form emails through the takeactionTN campaign promoted by the Tennessee Electric Cooperative Association and America’s Electric Cooperatives. These emails advocated an “all-of-the-above” approach to energy generation, opposed greenhouse gas regulations proposed by the EPA, expressed concern over reliance on nuclear and natural gas generation and emphasized low cost and reliability.

3.2 Public Involvement in Developing Study Inputs and Framework

IRP Process Several comments addressed aspects of the integrated resource planning process. Many of these supported the use of least-cost analysis and asked TVA to be sensitive to the adopted plan’s impact on ratepayers.

In this step, we used the key themes and results identified from public scoping to help develop the study framework, including the range of strategies for IRP analysis. During this phase, we worked closely with the IRP Working Group and held public briefings.

Comments on planning scenarios included the incorporation of the effects of climate change; varying approaches to incorporating regulation of greenhouse gas emissions; the evaluation of future fuel prices, particularly for natural gas; and the impacts of current and anticipated environmental regulations.

3.2.1 IRP Working Group Meetings Beginning in November 2013, TVA met with the IRP Working Group approximately every month. Twelve meetings were held prior to the release of the Draft IRP and associated SEIS at various locations throughout the region. Two additional meetings were held during and after the closure of the public comment period.

Comments on resource planning strategies included maximizing renewable generation and energy efficiency, phasing out the use of fossil fuels, transmission grid upgrades and increasing distributed generation.

The meetings were designed to encourage discussion on all facets of the process and to facilitate information sharing, collaboration and expectation setting for the IRP. IRP Working Group members reviewed and commented on planning assumptions, analytical techniques and proposed energy resource options and strategies. Specific topics included the energy efficiency approach in the IRP models, TVA’s power delivery

Other comments regarding the planning process addressed the valuation of renewable energy resources, the removal of limits on quantities of renewable energy, energy efficiency, demand response, and incorporating the external health and environmental costs of all energy resources.

18



structure, load and commodity forecasts and supply resource options. Given the diverse makeup of the IRP Working Group, there was a wide range of views on specific issues, such as the value of energy efficiency programs, environmental concerns and the costs associated with various generation technologies. Open discussions supported by the best available data helped improve understanding of the specific issues.

Location

November 5, 2013 December 5, 2013 January 13, 2014 February 19, 2014 March 27 & 28, 2014 April 29 & 20, 2014 May 29 & 30, 2014 June 19 & 20, 2014 October 7, 2014 Dec 15 & 16, 2014 January 26 & 27, 2015 February 26, 2015 April 9 &10, 2015 June 16, 2015

Knoxville, Tenn. Knoxville, Tenn. Murfreesboro, Tenn. Chattanooga, Tenn. Chattanooga, Tenn. Knoxville, Tenn. Chattanooga, Tenn. Knoxville, Tenn. Chattanooga, Tenn. Knoxville, Tenn. Knoxville, Tenn. Webinar Huntsville, AL. Knoxville, Tenn.

Location

March 26, 2014 June 18, 2014 November 3, 2014

Chattanooga, Tenn. Knoxville, Tenn. Knoxville, Tenn.

Figure 3‑2: Inputs and Framework Public Briefings Participants had the option to attend in person or participate by webinar. At each meeting, TVA staff made a brief presentation, followed by a moderated Q&A session.

To increase public access to the IRP process, all nonconfidential IRP Working Group meeting material was posted on TVA’s website, along with webinar recordings and related presentation videos. We also developed an IRP Working Group website for members to post information and to request data from our staff. Date

Date

Topics discussed at the public briefings included an introduction to the integrated resource planning process, resource options, development of scenarios and strategies and evaluation metrics. An average of 20 people attended each of the first three public briefings in person, and approximately 50 people participated via webinar. Recordings of the sessions were posted on the IRP website. TVA also briefed the public on the IRP process through presentations to local organizations, clubs and associations.

3.3 Public Involvement in Review of the Draft IRP TVA provided the draft IRP for public comment from March 13, 2015, through April 27, 2015. During this time TVA also held public meetings around the region to provide an opportunity for residents and stakeholders to learn more about the draft IRP, ask questions and provide general feedback. These information and feedback opportunities are also consistent with our obligations under the National Environmental Protection Act (NEPA).

3.2.2 Public Briefings In addition to the public scoping and IRP Working Group meetings, TVA held three briefings to update the public on the status of IRP development. Figure 3-2 shows the dates and locations of these briefings.

Written comments were also accepted online, mail and email.

19



3.3.1 Public Meetings

• Many advocated earlier increased use of EE and renewable energy (during first 5 years) and for a larger reduction in fossil fuel generation. • Greenhouse gas emissions and associated climate change are the most frequently mentioned environmental concern. • The majority of detailed / technical comments are on modeling data inputs and assumptions. • Several comments point out that given the very small differences in most strategy metrics, TVA should select a strategy with high levels of EE and renewable energy given the differences in environmental metrics and TVA’s environmental stewardship mission. • Several comments emphasized the importance of power cost, reliability, and availability.

Seven public meetings were held around the TVA region during the public comment period. Date

Location

March 19, 2015 April 6, 2015 April 9, 2015 April 14, 2015 April 15, 2015 April 21, 2015 April 22, 2015

Chattanooga, Tenn.* Knoxville, Tenn.* Huntsville, AL Tupelo, Miss. Memphis, Tenn. Nashville, Tenn. Bowling Green, Ky.

* Also Webinar / Live Streamed Meeting

Figure 3‑3: Draft IRP Public Meetings

The approximately 125 unique comments submitted can be grouped into the following themes:

At each of these meetings, TVA presented an overview of the Draft IRP, followed by a moderated Q&A session supported by a panel of TVA subject-matter experts. Attendees were able to address comments or questions to the panel. Attendees also had the option to submit written comments or online comments at the meetings. Written and online comments were also accepted during the full public comment period in addition to at the meetings. Approximately 400 people attended the public meetings in person and on-line.

• Alignment with TVA’s mission and least-cost planning mandate • Strategic direction and the final portfolio • Solar modeling • Modeling of wind options • Treatment of energy efficiency as a resource • Strategy evaluation metrics (scorecard) • Stakeholder engagement process

Verbal and specific written comments enabled TVA staff to identify public concerns and recommendations concerning the future operation of the TVA power system.

Specific responses to these comments are provided in Volume 2 of the SEIS. In this section we have summarized TVA’s general response to each of these broader themes:

3.4 Public Comment General Themes

1. Alignment with TVA’s mission and least-cost planning mandate Comment Issue: The current IRP does not properly fulfill the mission because it does not do least-cost planning properly with regard to modeling/selection of energy efficiency. On the other hand, some methodology improvements have enhanced TVA’s ability to do leastcost resource planning.

Reviewing comments received during the comment period yielded the following general observations: • There is widespread support for the overall IRP development process, including the approach to modeling energy efficiency and renewable energy as selectable resources and our extensive public involvement. • A large majority of comments advocated increased use of EE and renewable energy.

Response: The IRP study was conducted consistent with TVA’s least-cost system planning mandate as contained in the TVA Act. The improvements we made

20



in our modeling methodology allowed TVA to analyze energy efficiency and renewable energy as resources that the model could select. This treats these resources consistent with the way more traditional energy resources are modeled. In doing this, we also took into account necessary features of system operation such as resource diversity, reliability, resource dispatchability and risk factors.

advocates in developing modeling inputs, and subjected all assumptions to third party review and validation against industry references. The uncertainty ranges around the cost of solar used in the study encompass recent market trends. Sensitivity cases were also conducted to assess the impact of changes in these assumptions. 4. Modeling of wind options Comment Issue: Wind was not properly modeled; especially the assumptions around NDC (Net Dependable Capacity) and capacity factors; and the characteristics of High Voltage Direct Current (HVDC) wind were not consistent with the TVRIX recommendations. A learning curve (technology improvement) should be included.

2. Strategic direction and the final portfolio Comment Issue: The IRP points in the right direction by demanding higher levels of alternative resources. All strategies are too close to declare a “winner” so final portfolio should have higher levels of EE and renewables for broader benefits to Valley residents. Response: We appreciate that most people commenting on the IRP think its direction is correct. The purpose of the IRP is to provide directional guidance to TVA and the TVA Board. If this guidance is approved by the TVA Board, TVA would have the flexibility to increase levels of energy efficiency and renewable resources. We agree that all of the strategies were fairly close across IRP metrics, but the differences are important and should not be discounted. For example, all of the IRP strategies reduce the environmental impacts of the TVA system, but the higher levels of energy efficiency and renewables in Strategies D and E have the best environmental results. Risks are minimized by maintaining a diverse portfolio which all of the strategies promote, but strategies that emphasize specific resources more (Strategies D and E) have higher risks. System flexibility is generally equivalent across strategies, but is lower when renewable energy is emphasized. None of these differences are trivial.

Response: Assumptions for wind options were accepted from advocates and subjected to third party validation. TVA elected to base modeling assumptions on current technology and reasonable assumptions for capacity value on peak and overall energy production. Sensitivity cases were also conducted to assess the impact of these assumptions on model results. TVA will continue to refine its modeling of wind resources. 5. Treating energy efficiency as a resource Comment Issue: Energy efficiency (EE) design parameters and risk assumptions are not well modeled. The IRP doesn’t take into account all the EE benefits (line losses, avoided transmission & distribution investment). The current model design makes EE appear too costly and risky. Current assumptions constrain deployment. Response: TVA modeled EE as a selectable resource, consistent with how other energy resources are traditionally modeled. This is innovative and was widely endorsed by commenters. To do this, TVA had to create new techniques to capture risk and cost trends over time. While our modeling of EE for this IRP represents a significant improvement over how EE traditionally is evaluated in IRP processes, we recognize that the methodology could still be improved, including accounting for the potential indirect benefits identified by the commenters. TVA will continue to

3. Solar modeling Comment Issue: The study did not properly include solar PV (rooftop) as an option. Initial cost assumptions are too high based on recent market pricing. Costs for this option should decline at a faster rate. Response: TVA captured impacts from distributed solar in the design of the scenarios; in addition, small commercial solar was included as an option in the cases. We considered assumptions from solar

21



refine its modeling methodology. We do think that EE is inherently more risky than other energy resources because TVA does not directly serve most end users of the electricity it generates. Local power companies have that relationship. To be successful, most energy efficiency programs have to be embraced by these end users and the LPC-end user interface adds another level of uncertainty. Sensitivity cases were conducted to test several of the key assumptions and results are generally consistent with the original findings. In nearly every case, additional deliveries from EE are envisioned.

the public were afforded multiple opportunities to provide feedback. Sensitive data, such as detailed cost data, had to be handled more carefully because of the potential harm to TVA’s business transactions. However, TVA even opened up this kind of sensitive data to stakeholder input and review. It was shared with members of the IRP Working Group on a confidential basis. TVA created sensitivity cases to test specific parameters in order to better understand the robustness of the IRP strategies. Many of these cases were created in response to comments that TVA received during the IRP process, particularly from members of the IRP Working Group. By necessity, this had to be done at the end of the planning effort. Sensitivity cases were fully discussed with members of the working group and the input we received was considered in the preparation of the final report. The final report contains information about the sensitivity cases.

6. Strategy evaluation metrics (scorecard) Comment Issue: Cost and risk are reasonable metrics, but risk is over-estimated for EE and renewables. The system average cost metric should be eliminated. Flexibility metric does not provide any insight of value; environmental metrics are not given sufficient weight in the assessment. Response: TVA developed the scorecard metrics with input from stakeholders to capture key attributes of the planning cases. We believe all the metrics have value and inform the decision around selecting a preferred planning direction. Average system cost provides a key signal about rate pressure. The flexibility metric provides an indication of general responsiveness of the system and is a first attempt at acknowledging and reporting this important metric. Stakeholders specifically requested we not weight any of the scorecard metrics so that findings would not be unduly shaped by the assigned weights.

IRP Working Group Jack Barkenbus, Senior Researcher Climate Change Research Network Vanderbilt University Nashville, TN Lance Brown, Executive Director Partnership for Affordable Clean Energy Montgomery, AL Donald (Chip) Cherry, President and CEO Huntsville Madison County Chamber of Commerce Huntsville, AL

7. Stakeholder engagement process Comment Issue: Transparency and engagement efforts were good. Recommend sharing of detailed cost data with the public. Some resource modeling decisions made without stakeholder input. Consider providing sensitivity case results for public comment.

Mary English, Senior Fellow University of Tennessee Baker Center for Public Policy Knoxville, TN Zachary Fabish, Staff Attorney Sierra Club Washington, DC

Response: TVA has uniformly received high marks by commenters for the level of stakeholder engagement during this IRP process. Several commenters pointed out that this level of engagement is not seen in other IRP processes. Stakeholder input was considered in virtually all phases of the study. Stakeholders and

Amanda Garcia, Staff Attorney Southern Environmental Law Center Nashville, TN

22



Keith Hayward, General Manager and CEO North East Mississippi Electric Power Association Oxford, MS

Jack Simmons, President and CEO Tennessee Valley Public Power Association Chattanooga, TN

Richard Holland, Vice President Tennessee Paper Council Nashville, TN

Jay Stowe, Chief Executive Officer Huntsville Utilities Huntsville, AL

Don Huffman, Executive Director Associated Valley Industries Chattanooga, TN

Michelle Walker, Assistant Commissioner Office of Policy and Planning Tennessee Department of Environment and Conservation Nashville, TN

Dana Jeanes, Vice President and Chief Financial Officer Memphis Light, Gas and Water Memphis, TN

Lloyd Webb, Strategic Planning Chair Tennessee Valley Industrial Committee Cleveland, TN

Tom King, Director Energy Efficiency and Electricity Technologies Oak Ridge National Lab Oak Ridge, TN

John Wilson, Research Director Southern Alliance for Clean Energy Washington, DC

David Rumbarger, President and CEO Community Development Foundation of Tupelo Tupelo, MS Kate Shanks, Executive Director Office of Legislative and Intergovernmental Affairs, Energy and Environment Cabinet Commonwealth of Kentucky Frankfort, KY

23

Chapter 4

Need for Power Analysis Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

25

Re-evaluate

Recommend


Chapter 4: Need for Power Analysis

4 Need for Power Analysis

processes in each sector and expected changes in the stock and efficiency of equipment and appliances.

A primary purpose of this IRP is to accurately determine whether the energy resources TVA currently has available are sufficient to supply the power the Tennessee Valley region will need over the study period (2014-2033) and, if the anticipated demand exceeds the current supply, to estimate the capacity gap and determine what type and how much additional generating resources are needed.

For example, in the residential sector, energy usage was forecasted for space heating, air conditioning, water heating and several other uses after accounting for changes in efficiency over time, appliance saturation and replacement rates, growth in average home size and other factors. In the commercial sector, we gave similar attention to changes in efficiency, saturation and other variables in a number of categories, including lighting, cooling, refrigeration and space heating.

This chapter describes the four steps in the process used to make this determination: compute demand, determine reserve capacity needs, estimate supply and estimate the capacity gap.

Finally, working with TVA customer service representatives, we supplemented the historical data used in our modeling with industry analyses and feedback from our large, directly served customers regarding demand. This input helped us better predict the magnitude and timing of changes in load attributable to plant closures and expansions.

4.1 Estimate Demand The first step in forecasting future power needs is to estimate long-term growth in electricity sales and peak demand. Peak demand, or peak load, is the highest one-hour power requirement placed on the system. In order to reliably serve customers, TVA must have sufficient resources to meet the peak hour demand.

Key Forecast drivers Growth in Economic Activity At least annually, TVA produces a forecast of regional economic activity for budgeting and long-range planning purposes. These forecasts are developed from county-level economic forecasts in order to accurately model the prevailing economic conditions in the region.

The electricity sales and peak demand forecasts for this IRP were developed from individual, detailed forecasts of residential, commercial and industrial sales. We checked the historical accuracy of these forecasts to help ensure errors in data or methodology were quickly identified and resolved. We also generated a range of forecasts (high, expected, and low) to ensure that TVA’s plans do not depend on the accuracy of a single forecast.

Historically, the Tennessee Valley economy was more dependent on manufacturing than the economies of other regions. Industries such as pulp and paper, aluminum, steel and chemicals were drawn to the Valley because of the availability of natural resources, access to a skilled workforce and the supply of reliable and affordable electricity. However, manufacturing’s share of non-farm employment has steadily declined in the Valley, as it has across the region.

4.1.1 Load Forecasting Methodology To forecast future electricity demand, TVA uses statistical and mathematical models that link electricity sales to several key drivers. These include the growth in overall economic activity, the price of electricity, customer retention and the price of competing energy sources such as natural gas.

Our region is different from others in that manufacturing’s share of economic output in the Tennessee Valley actually increased since 1980, from 17 percent to 18 percent. These trends speak to the changing nature of economic activity here. While many labor-intensive manufacturing industries moved

We also apply an end-use forecasting model to capture the effect of underlying trends affecting residential, commercial and industrial electricity sales such as changes in the use of various types of equipment or

26



overseas, a continued shift toward energy-efficient manufacturing processes in the Valley is helping to preserve manufacturing’s contribution to total economic output. This is important to TVA’s load forecasting in that it may indicate a weakening in the historical relationships between economic growth and load growth.

plants outside TVA’s service area if TVA’s rates become non-competitive. Additionally, large industrial operations could generate their own power without distribution or transmission line losses – an increasingly attractive option to TVA’s largest customers as hydraulic fracturing reduces the cost of natural gas to unexpected lows. These risks are factored into TVA’s load forecasts because they could affect future load, but we believe they will be offset by our commitment to keeping TVA rates competitive.

Because of this continued dependence on manufacturing, our region’s economy tends to be more sensitive to economic conditions impacting the demand for manufactured goods. Near-term future growth in 2015 and 2016 is expected to benefit from positive cyclical economic conditions. After 2016, however, longer-term demographic pressures are expected to hold average growth in Gross Regional Product near 1.6 percent as retiring baby boomers restrict the available labor supply. Population growth in the Tennessee Valley declined from an annual average of about 1.0 percent in 1980 to 0.7 percent in 2014 and is expected to decline to 0.5 percent by 2043, which will limit the demand for all goods and services, including power.3

Price of Competing Energy Sources Changes in the price of electricity compared to the price of natural gas and other fuels also influences load growth.

Price of Electricity Forecasts of the retail price for electricity are based on long-term estimates of TVA’s total costs to operate and maintain the power system and are adjusted to include an estimate of the historical markups charged by distributors of TVA power. These costs, known in the industry as revenue requirements, are based on estimates of the key costs of generating and delivering electricity, including fuel, variable operations and maintenance costs, capital investment and interest.

If consumers can heat their homes and water cheaper using natural gas or other energy sources, they may move away from electricity in the long-term. The potential for this type of substitution depends on the relative prices of other fuels and the ability of those fuels to provide a comparable service. It also depends on the physical capability to make the substitution. For example, while consumers can change out electric water heaters and replace electric heat pumps with natural gas furnaces, the ability to use another form of energy to power consumer electronics, lighting and cooling is far more limited by current technology. Changes in the price of TVA electricity compared to the price of natural gas and other fuels also influence consumers’ choices of appliances—either electric, gas or other fuels.

Customer Retention Over the past 25 years, the electric utility industry has undergone a fundamental change in most parts of the country. In many states an environment of regulated monopoly has been replaced with varying degrees of competition. Although TVA has contracts with the 155 local power companies (LPCs), it is not immune to competitive pressures. The contracts allow LPCs to give TVA notice of contract cancellation after which they may buy power from other sources. Many large industrial customers also have the option of shifting production to

While other substitutions are possible, the price of natural gas serves as the benchmark for the relative competitiveness of electricity and the potential substitution impacts on load forecasts. Accounting for the long-run impact of natural gas prices is especially important in light of the increased competitive pressure resulting from hydraulic fracturing and the shale gas revolution. Although low gas prices make power production less expensive, it also tempts customers to shift from electricity to natural gas to meet their energy needs.

3 TVA population data from U.S. Census Bureau and Moody’s Analytics

27



4.1.2 Forecast Accuracy

forecast, and the year-ahead forecast variances tend to be smaller.

Forecast accuracy is generally measured by how much the forecast deviates from the actual energy and peak demands, adjusted for abnormal weather. Figures 4-1 and 4-2 show annual forecasts for fiscal years (FY) 1999 through 2014 for peak load requirements and net system energy requirements compared to actual peak loads and actual energy use. Figure 4-1 is a comparison of actual annual peak demands in megawatts (MW) to the peak demands forecasted one year earlier. The red “Normalized Actual” line represents what the annual peak would have been under normal weather conditions. The closer the blue dotted “Forecast” line is to the red “Normalized Actual” line, the more accurate the peak forecast. For example, in FY14, the actual peak was only 1.3 percent greater than forecasted. Over-forecasts from FY11 to FY13 are related to the Great Recession, which resulted in a decline in weather-normalized peaks that continued well after the recession’s end. We are now seeing the return of modest growth in weather-normalized peaks.

Figure 4‑2: Comparison of Actual and Forecasted Net System Requirements The mean absolute percent error (MAPE)4 of TVA’s forecasts of net system energy and peak load requirements FY 1999 through FY 2014 was 1.8 percent and 3.1 percent, respectively. While the nature of forecasting a single hour’s MWs inherently leads to elevated peak forecast errors, this includes an unusually large error (7.6 percent) in the FY09 peak forecast as the full severity of the Great Recession’s impact was not yet fully realized. Energy is less volatile so forecast errors tend to be a little smaller, but the Great Recession still adversely impacted the energy forecast’s error rate. Ignoring the forecast for 2009 brings the energy error rate down from 1.8 to 1.4 percent, which is more in line with what we expect in a typical year. From informal conversations with peer utilities, TVA’s energy forecast MAPE of around 1 to 2 percent is in line with other utilities.

Figure 4‑1: Comparison of Actual and Forecasted Annual Peak Demand

4.1.3 Forecasts of Peak Load and Energy Requirements

Figure 4-2 is a comparison of actual and forecasted net system requirements expressed in total annual energy, measured in gigawatt-hours (GWh). Energy is somewhat less volatile than peaks, which are based on a single hour of each year, because energy is the sum of all the hours of the year. This makes energy a little easier to

4 MAPE is the average absolute value of the error each year; it does not allow overpredictions and under-predictions to cancel each other out. 5 Refer to Chapter 6 for a discussion of the scenarios developed for this IRP.

Over the next couple decades the Current Outlook Scenario5 anticipates net system energy and peak demand to grow about 1.0 percent and 1.1 respectively, which is somewhat slower than the 1.3 percent experienced for both net system energy and peak

28



demand over the FY1990 through FY2013 period. These lowered expectations from the long-term trend are a function of both economics and energy efficiency projections. Slower economic growth, driven by the baby boomers’ retirement, and an ever-tightening regulatory environment are both anticipated to moderate future energy growth.

The use of ranges ensures that TVA considers a spectrum of electricity demand in its service territory and reduces the likelihood that its plans are overly dependent on a single-point estimate of demand growth. Alternative scenarios highlight the risk inherent in forecasting and planning to a single point estimate. The scenario-generated ranges are used to inform planning decisions beyond pure least-cost considerations based on a specific demand in each year.

To deal with the inherent uncertainty in forecasting, TVA has developed a range of forecasts. Each forecast corresponds to different load scenarios around the Current Outlook Scenario’s forecast. The Current Outlook Scenario for the IRP is the forecast that TVA prepared for the FY2015 Long Range Financial Plan in fall of 2013. The range of forecasts for net system peak load and energy requirements used in the IRP are shown in Figures 4-3 and 4-4, respectively. Both include the Current Outlook Scenario and the highest and lowest growth scenarios that were modeled. They are the Growth Economy Scenario and the Distributed Marketplace Scenario, respectively. Annual peak load growth over the 2014 through 2033 time period varies from 0.3 percent in the lowest growth scenario to 1.3 percent in the highest growth scenario. Net system energy requirements grow at an annual rate of 1.0 percent in the Current Outlook Scenario but growth dips as low as 0.0 percent in the lowest growth scenario and peaks at 1.1 percent in the highest growth scenario.

Figure 4‑4: Energy Forecast

4.2 Determine Reserve Capacity Needs To maintain reliability, power providers must always have more generating capacity available than required to meet peak demand. This additional generation, called “reserve capacity,” must be large enough to cover the loss of the largest single operating unit (contingency reserves), be able to respond to moment-by-moment changes in system load (regulating reserves) and replace contingency resources should they fail (replacement reserves). Total reserves must also be sufficient to cover uncertainties such as unplanned unit outages, undelivered purchased capacity, severe weather events or load forecasting error. TVA identified a planning reserve margin based on minimizing overall cost of reliability to the customer. This reserve margin is based on a probabilistic analysis that considered the uncertainty of unit availability,

Figure 4‑3: Peak Demand Forecast

29



transmission capability, weather-dependent unit capabilities (e.g., hydro, wind and solar), economic growth and weather variations to compute expected reliability costs. Based on this analysis, we selected a target reserve margin that minimized the cost of additional reserves plus the cost of reliability events to the customer. The target or optimal reserve margin was adjusted based on TVA’s risk tolerance. Based on this methodology, TVA’s current planning reserve margin is 15 percent above peak load requirements and is applied during both the summer and winter seasons.

limited period even though it may be less economical to do so. This may be due to transmission or other power system constraints. Similarly, some baseload units are capable of operating at different power levels, giving them some characteristics of an intermediate or peaking unit. This IRP considered strategies that take advantage of this range of operations.

4.3 Estimate Supply The third step in the process of analyzing future power needs is to identify the supply- and demandside resources currently available to meet future power demand. Our generation supply consists of a combination of existing TVA-owned resources; budgeted and approved projects such as new plant additions, updates to existing assets; and existing power purchase agreements (PPAs). Generating assets can be categorized both by whether the power they produce is used to meet base, intermediate or peak demand or used for storage, and by capacity type or energy/fuel source.

Figure 4‑5: Illustration of Baseload, Intermediate and Peaking Resources

4.3.1.1 Baseload Resources

4.3.1 Baseload, Intermediate, Peaking and Storage Resources

Due to their lower operating costs and high availability, baseload resources are used primarily to provide continuous, reliable power over long periods of uniform demand. Baseload resources typically have higher construction costs than other alternatives, but also have much lower fuel and variable costs, especially when fixed costs are expressed on a unit basis (e.g., dollars per MWh). An example of a baseload resource is a nuclear power plant.

Figure 4-5 illustrates the uses of baseload, intermediate and peaking assets. Although these categories are useful, the distinction Power purchase agreements between them is not (PPAs) refer to the energy always clear-cut. For and/or capacity bought from example, a peaking other suppliers for use on the unit, which is typically TVA system in place of TVA used to serve only building and operating its own intermittent but resources. Power purchases short-lived spikes in provide additional diversity for demand, may be TVA’s portfolio. We are currently called on from time to a party to numerous short-term time to run and long-term PPAs. continuously for a

Some energy providers also use natural gas-fired combined cycle (CC) plants as incremental baseload generators. However, given the historical tendency for natural gas prices to be higher than coal and nuclear fuel prices when expressed on a unit basis (i.e., dollar per million British Thermal Unit), a CC unit is generally considered a more expensive option for larger

30



continuous generation needs. Although natural gas-fired CC plants could become more attractive for baseload generation as the fundamentals of fuel supply and demand change and if access to shale gas continues to grow.

function as peaking units but use low-cost, off-peak electricity to store energy for generation at peak times. An example of a storage unit is a hydroelectric pumped-storage plant. These plants pump water to a reservoir during periods of low demand and release it to generate electricity during periods of high demand. Consequently, a storage unit is both a power supply source and an electricity user.

4.3.1.2 Intermediate Resources Intermediate resources are used primarily to fill the gap in generation between baseload and peaking needs. They also provide back-up and balance the supply of energy from intermittent wind and solar generation.

4.3.2 Capacity and Energy Power system peaks are measured in terms of capacity, the instantaneous maximum amount of energy that can be supplied by a generating plant and collectively by the power system.

Intermediate units are required to produce more or less output as the energy demand increases and decreases over time, both during the course of a day and seasonally. Given current fuel prices and relative generating efficiencies, these units are more costly to operate than baseload units but cheaper than peaking units.

For long term planning purposes, capacity can be defined in several ways: • Nameplate capacity is the theoretical design value or intended maximum megawatt output of a generator at the time of installation. • Capability is the maximum dependable loadcarrying ability of units or the number of megawatts that can be delivered by a generating unit without restrictions (i.e., does not reflect temporary capacity restrictions caused by known fuel or mechanical derates) and less station power. • Net dependable capacity is the maximum dependable output less all known adjustments (e.g., transmission restrictions, station service needs and fuel derates) and is dependent on the season. This value, which is used by capacity planners, is typically determined by performance testing during the respective season. TVA uses the summer net dependable capacity of a unit because the capacity of thermal generating units is reduced during the heat of summer which is when the load on the TVA system typically peaks.

Intermediate generation typically comes from natural gas-fired CC plants and smaller coal units. However, energy from wind and solar generation also can be used as intermediate resources depending on the energy production profile and the availability of energy storage technologies. Hydro generating assets can generally be categorized as intermediate resources, but their flexibility allows them to operate the full range from baseload to peaking. The limitation of hydro generation is restricted more by water availability and the various needs of the river system such as navigation.

4.3.1.3 Peaking Resources Peaking units are expected to operate infrequently during short duration, high demand periods. They are essential for maintaining system reliability requirements, as they can ramp up quickly to meet sudden changes in either demand or supply. Typical peaking resources are natural gas-fired combustion turbines (CTs), conventional hydroelectric generation and pumpedstorage hydroelectric generation.

Overall power system production is measured in terms of energy (i.e., megawatt-hour). Energy is the total amount of power that an asset delivers in a specified time frame. For example, one MW of power delivered for one hour equals one megawatt-hour (MWh) of energy.

4.3.1.4 Storage Resources Storage units usually serve the same power supply

The capacity factor of a power plant is a measure of

31



the actual energy delivered by a generator compared to the maximum amount it could have produced at the nameplate capacity. Assets that run constantly, such as nuclear plants, provide a significant amount of energy with capacity factors greater than 90 percent. Capacity Factor Examples High capacity factor unit: A 1,200 MW unit could theoretically produce 10,512 GWh of power if it ran every hour of the year. After planned annual outages, the unit will typically produce 9,461 GWh or 90 percent of its theoretical capacity. Low capacity factor unit: A 250 MW natural gas-fired combustion turbine (CT) unit could theoretically produce 2,190 GWh of power if it ran every hour of the year. However, CT units generally have a capacity factor less than 5%, which means the unit would likely operate about 438 hours of the year and produce 110 GWh.

and renewables6. This resource plan will be used in the associated EIS report to represent the “no action alternative” as required under NEPA. Figure 4-6 includes both owned and purchased resources, in megawatts of summer net dependable capacity, and is divided into fuel-type (i.e., nuclear, hydro, coal, etc.). In this chart, the lower area in the figure contains the existing assets and the new expansion assets are shown at the upper portion of the stacked area chart. EE is shown in the upper area of this chart to make comparison to the results from the modeling discussed in Chapter 7 easier. There are existing contracts for demand response that need to be accounted for and are shown in the existing assets. Similarly, the current renewable resources are shown in the bottom of the chart.

Assets that are used infrequently, such as a combustion turbine, provide relatively little energy with capacity factors of less than five percent, although the energy they produce is crucial since it is often delivered at peak times.

Figure 4-6 shows how TVA’s existing capacity portfolio is expected to change through 2033. The existing assets only include resources that currently exist; assets that are under contract; TVA Board-approved changes to existing resources such as refurbishment projects; and TVA Board-approved additions such as Watts Bar Unit 2. Existing resources decrease through 2033 primarily because of the retirement of coal-fired units and the expiration of existing contracts (power purchase agreements). The renewable component of the existing portfolio is primarily composed of wind PPAs. Because the power generated from wind and other renewable resources is intermittent, the firm capacity (or the amount of capacity that can be applied to firm requirements) for these assets is much lower than the nameplate capacity.

Energy efficiency also can be measured in terms of capacity and energy. Even though energy efficiency does not input power into the system, the effect is similar because it represents power that is not required from another resource. Demand reduction also is measured in capacity and energy. However, unlike energy efficiency, it does not offer a significant reduction in total energy used.

Having a diverse portfolio of resource types – coal, nuclear, hydroelectric, natural gas and renewable resources – and being able to use these resources in different ways enables TVA to provide reliable, low-cost power while minimizing the risk of disproportionate reliance on any one type of resource.

4.3.3 Current TVA Capacity and Energy Supply TVA uses a wide range of technologies to meet the needs of Tennessee Valley residents, businesses and industries. Figure 4-6 shows the generating assets that would be used to meet those needs over time for the baseline case, which reflects our current practice of optimizing the resource portfolio while scheduling the contribution of energy efficiency, demand response,

6 This IRP uses a new methodology to dynamically select these resources. See discussion in Chapter 6.

32



Figure 4-7 shows TVA’s estimated capacity gap or shortfall based on the existing firm capacity and the annual firm requirement for the current outlook scenario.

Capacity for the Baseline Case

MW,SND 45,000 40,000

New DR New EE

35,000

Capacity Gap

New Gas CC New Gas CT

30,000

New Hydro 25,000

New Coal

MW, SND

New Nuclear

20,000

45,000

Existing DR Existing Gas CC

15,000

Firm Requirement = Peak Demand + 15% Reserves - Interruptibles

40,000

Existing Gas CT/Diesel 10,000

Existing Coal

5,000 0 2014

30,000

Existing Hydro Existing Nuclear 2016

2018

2020

2022

2024

2026

2028

2030

Capacity Gap

35,000

Existing Renewables

25,000

2032 20,000

Figure 4‑6: Baseline Capacity, Summer Net Dependable MW

Existing Firm Suply

15,000 10,000 5,000

Approximately 37 percent of TVA’s capacity is currently sourced from emission-free assets such as nuclear power, renewable resources including hydroelectricity, and interruptible load management. The renewable category shown throughout this document is based on modeled outputs of energy from renewable sources such as wind, solar, and biomass. Therefore, this metric is not intended to represent a quantity of certified renewable energy credits.

-

2014

2016

2018

2020

2022

2024

2026

2028

2030

2032

Figure 4‑7: Estimating the Capacity Gap Figure 4-8 shows the range of capacity gaps corresponding to the highest growth scenario, the Economic Growth scenario, and the lowest growth scenario, the Distributed Marketplace scenario. These and other scenarios are described in detail in Chapter 6.

Currently, 33 percent of TVA’s electricity is produced from the nuclear fleet. Coal produces about 40 percent of the generation, hydroelectric plants produce approximately 10 percent, and 3 percent was produced from renewable sources. The gas fleet produces about 13 percent with the majority of that generation being provided by combined cycle plants and the remaining generation results from energy efficiency efforts.

Capacity Gap Range

4.4 Calculate the Capacity Gap The need for power can be expressed either as a capacity gap or as an energy gap. As noted previously, a capacity gap is the difference between total supply and total demand. More specifically, it is the difference in megawatts between a power provider’s existing firm capacity and the forecast annual peak adjusted for any interruptible customer loads plus 15 percent reserve requirements.

Figure 4‑8: Capacity Gap Range An energy gap is the amount of energy specified in GWh provided by the existing firm capacity resources minus the energy required to meet net system

33



requirements (i.e., the energy needed to serve the load over the entire year). It includes the energy consumed by the end-users plus distribution and transmission losses. Figure 4-9 shows the range of energy gaps TVA can expect under the net system requirements associated with the highest and lowest growth scenarios. Energy Gap Range

Figure 4‑9: Energy Gap

34

Chapter 5

Energy Resource Options Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

35

Re-evaluate

Recommend


Chapter 5: Energy Resource Options

5 Energy Resource Options

the agency is called upon to be a leader in Technology Innovation. Currently, we have identified three signature technologies for special emphasis: small modular reactors, grid modernization and energy utilization focusing on efficiency and demand response products. In support of our broader technology innovation efforts, we have included small modular reactors as an option in this IRP; and the technology advancements around energy utilization are also included in this study as part of our modeling of EE and DR as selectable resource options.

Maintaining the diversity of TVA’s energy resources is fundamental to our ability to provide low-cost, reliable and clean electric power to Valley residents, businesses and industries. For this reason, we considered the addition of a wide range of supply-side generating resources, as well as energy efficiency and other demand-side resource options, to fill the forecasted 20-year capacity and energy gaps identified through the power needs analysis described in Chapter 4. The power needs analysis indicates that, under the Current Outlook scenario, TVA will require additional capacity and energy of 2,100 MW and almost 16,000 GWh by 2020, growing to 10,300 MW and more than 58,000 GWh by 2033.

5.1.2 Characteristics Required for Resource Options To compare energy resource options available for new generation fairly, it is important to have consistent data regarding the cost and operating characteristics of each option. A list of characteristics used in the 2015 IRP are identified and defined below. Section 5.2.2 will present the numerical values for some of these parameters for the new assets.

5.1 Energy Resource Selection Criteria During the scoping process, TVA identified a broad range of energy resources that could be used to fill the predicted capacity and energy gaps. The next two sections explain the criteria that were used to reduce this list to a manageable portfolio of expansion options. For a complete list of resource options considered, see Chapter 5, Energy Resource Options, of the associated EIS.

Cost characteristics: • Unit capital costs: Each technology type must have a representative $/kW, which is considered a total installed cost. Total installed cost includes equipment, engineering and interest during construction in present day dollars. • Capital escalation rates: Since capital costs typically increase over time, a simplifying assumption could be that the capital costs escalate at the forecast rate of inflation. However, some renewable energy technologies are forecast to decrease over time. • Construction spend schedule: Some technologies take a long time to build. Construction times for nuclear units, for example, average about 10 years. To estimate the cash flow for the construction of a long-lead time build unit such as a nuclear unit, the percent of total capital dollars spent in each year is required. This metric is typically not needed for renewable assets which are smaller in scale and generally built in less than a year. • Fixed operating and maintenance costs (FOM): FOM costs are independent of the number of hours of operation or amount of electricity produced and are

5.1.1 Criteria for Considering Resource Options Two criteria were used to ensure that only viable energy resource options were considered in the IRP analysis. To be considered, resource options must: Use a proven technology, or one that has reasonable prospects of becoming commercially available in the planning horizon Be available to TVA within the region or be available to be imported through market purchases Technology is a key factor in TVA’s ability to fulfill its mission in a balanced way. TVA continues to pursue technological advances to become more efficient and sustainable. As part of its mission under the TVA Act,

36



generally expressed in a dollar per kilowatt per year ($/kW-yr). FOM includes operating and maintenance labor, plant support equipment, administrative expenses and fees required by regulatory bodies. • Variable operating and maintenance costs (VOM): VOM costs are dependent on the number of hours of operation and are generally expressed as a dollar per megawatt-hour ($/MWh). VOM costs include consumables like raw water, waste and water disposal expenses, chemicals and reagents. VOM costs do not include fuel expenses. • Fuel expenses: Fuel is the material that is consumed to generate electricity – for example, coal, natural gas, uranium and biomass. These costs are typically expressed in a dollar per million British thermal units ($/mmBtu) and include the delivery charges. • Transmission: A new generating resource has to be connected to the transmission system. Costs are typically expressed as a dollar per kilowatt ($/ kilowatt).

commercial availability, as well as permitting and construction times. For example, if it takes four years to build a combined cycle plant, then a new CC could not be selected prior to four years into the planning horizon. • Book life: The book life of a unit is the number of years a resource is expected to be in service for accounting purposes. Book life is the financial payback period which represents the amount of time the asset is expected to be used and useful. A license extension, beyond the original asset life, is not assumed with any new generating option.

5.2 Resource Options Included in IRP Evaluation TVA’s existing assets, including existing TVA-owned resources, as well as budgeted and approved projects, and power purchase agreements, are considered fixed assets in the IRP evaluation. These assets are expected to continue operating through the duration of the planning period or through the terms of existing power purchase agreements and other contracts, where applicable.

Operating characteristics: • Summer net dependable capacity: Each unit must have a summer net dependable capacity rating in megawatts. • Capacity credit: The capacity credit must be estimated for variable units or non-dispatchable resources. The capacity credit is the amount of capacity immediately available at the highest demand times. • Summer full load heat rate: A heat rate must be specified for each unit. A heat rate is a measure of the consumption of fuel necessary for a unit to produce electricity. Heat rates are shown in British thermal units per kilowatt hour (Btu/kWh) and are based on a summer full-load heat rate. Heat rates are considered long-term planning assumptions and include the expected degradation in the heat rate of a unit after the first two years. Although a heat rate is not typically associated with a nuclear unit, one is necessary to model the fuel costs. • Unit availability: A date when each unit would be available for operation must be specified. Unit availability is restricted by technical feasibility or

Options for new generation to meet the forecast net system requirements identified in Chapter 4 include: building new generating units, retro-fitting existing units with controls to continue operations, development of energy efficiency and demand response programs, and new power purchase agreements. The next two sections describe existing and potential new generation by resource category. For a comprehensive description of all resource option attributes, characteristics and technologies, see Chapter 5, Energy Resource Options, of the associated EIS.

5.2.1 Existing Assets by Resource Category Nuclear TVA currently operates six nuclear reactors: three at Browns Ferry Nuclear Plant, two at Sequoyah Nuclear

37



Plant and one at Watts Bar Nuclear Plant. These plants have a combined generating capacity of about 6,700 MW. On August 1, 2007, the TVA Board of Directors approved the completion of a second reactor at Watts Bar Nuclear Plant. This reactor will have a 1,150 MW generating capacity. The new reactor is scheduled to become operational by the end of 2015 and is included as a current resource in TVA’s generating portfolio.

a generating system to carry power for a specified time and does not include operational limitations such as fuel de-rates. We use a value lower than the capability of a resource for the summer net dependable capacity. By 2016, the existing coal fleet will decrease to about 35 active units with a total capability of 10,300 MW. Below is a snapshot of the planning assumptions for the coal units.

Coal TVA operates 10 coal-fired power plants consisting of 41 active generating units with a total capability of almost 11,900 MW. Capability is defined as the ability of

In addition to TVA-owned coaled fired units, TVA has access to the output from a coal-fired power plant with a generating capacity of about 440 MW through a longterm power purchase agreement.

Figure 5‑1: Coal Fleet Map 38



Coal Plant

Total # of Original Units

Current Operating Status

Operational Plan

Allen

3

Operational

Retire all three units. Board approved plan to construct a 2x1 combined cycle plant adjacent to the site

Bull Run

1

Operational

Continue to operate

Colbert

5

Unit 5 idled Units 1-4 operational

Board approved plan to retire all five units

Cumberland

2

Operational

Continue to operate

Gallatin

4

Operational

Continue to operate with Board approved scrubbers and SCRs

Johnsonville

10

Units 1-4 operational Units 5-10 idled

Retire all units

Kingston

9

Operational

Continue to operate

Paradise

3

Operational

Board approved plans to construct a combined cycle plant on site, retire units 1 and 2, and continue operation of unit 3

Shawnee

10

Units 1-9 operational Unit 10 retired

Board approved plans to control units 1 and 4. The remaining units will continue to operate until a long-term decision is made

Widow’s Creek

8

Units 1-6 retired Unit 8 idled Unit 7 operational

Board approved plan to retire all units

Figure 5‑2: Coal Fleet Portfolio Plans Natural Gas TVA operates 87 combustion turbines (CT) at nine power plants with a combined generating capability of about 5,400 MW and 11 combined cycle (CC) units at five plants with approximately 3,900 MW of capability. TVA is also currently a party to a long-term lease of a 700 MW CC plant.

Hydroelectricity TVA operates 109 conventional hydroelectric generating units at 29 dams. These units have the capability to generate about 3,800 MW of electricity. In addition, TVA has a long-term power purchase agreement with the U.S. Army Corps of Engineers for eight dams on the Cumberland River system. These facilities provide almost 400 MW of capability.

Petroleum Fuels TVA currently owns five diesel generators and has a few other diesel generators under power purchase contracts. These resources provide a total capability of about 120 MW.

TVA anticipates about 60 percent of the capability to be available at the summer peak hour given all the operational constraints.

39



Energy Storage TVA operates one large energy storage facility. Our Raccoon Mountain Pumped-Storage Plant has four generating units with a SND capacity of 1,616 megawatts. Raccoon Mountain is TVA’s largest hydroelectric facility and provides critical flexibility to the TVA system by storing water at off-peak times for use when demand is high.

Energy Efficiency TVA’s energy efficiency portfolio focuses on reduction in peak demand and energy savings. From FY2007FY2014 these efforts contributed 535 MW of summer peak demand reduction and save 1,506 GWh of energy annually. Impacts are realized at the generator and include applicable transmission and distribution (T&D) losses, free rider/driver discounts, realization rates, and performance adjustments for actual weather. This differs from other sources, such as the ERS Highlight’s Report, which do not normally include these factors; and are more reflective of end-user savings.

Wind TVA purchases all of the power produced by the Buffalo Mountain wind farm in Anderson County, Tenn. Buffalo Mountain is the largest wind farm in the Southeast, with 18 turbines and 27 MW of nameplate capacity. As defined in section 4.3.2, the nameplate capacity is the maximum technical output of a generator, or the theoretical design value.

Demand Response Demand response programs also focus on reduction of peak demand. Under these programs, TVA industrial and commercial customers can reduce their power bills by allowing TVA to suspend availability of power in the event of a power system emergency. These programs provide about 600 MWs of peak reduction. Another program allows TVA to curtail power delivery to participants for economic or reliability reasons. This program provides about 560 MWs of peak reduction. If needed, TVA also can reduce peak demand by about 85 MWs through in-house curtailments.

We also have long-term power purchase contracts with eight wind farms located in Illinois, Kansas and Iowa. These facilities provide about 1,500 MW of nameplate capacity. TVA anticipates about 14 percent of the nameplate to be available for peak summer requirements. TVA obtains the renewable energy credits from seven of these farms. Renewable energy credits are a separate commodity formed from the production of energy at designated sites.

5.2.2 New Assets by Resource Category A complete list of viable new resource options for IRP evaluation is provided below. A detailed discussion by resource category follows.

Solar TVA owns 16 photovoltaic (PV) installations with a combined capacity of about 300 kilowatts of nameplate capacity. We also purchase solar power through several programs and long-term power contracts totaling nearly 72 MW of nameplate capacity with about 36 MW expected to be available at the summer peak hour.

An independent third-party reviewed and compared the parameters to proprietary and other industry sources to ensure the modeled unit characteristics and assumptions were representative of the respective generating technologies. See Appendix A for the letter summary of the benchmarking efforts of Navigant Consulting, Inc. as well as a brief discussion of TVA’s internal benchmarking on resource costs ($/kW).

Biomass TVA generates electricity at Allen Fossil Plant by cofiring methane from a nearby sewage treatment plant and by co-firing wood waste at Colbert Fossil Plant. The co-firing is more like a fuel switch for coal and does not provide additional capacity to either of the coal plants. TVA purchases about 49 MW of biomass-fueled generation.

40



Nuclear • Pressurized water reactor (PWR) • Advanced pressurized water reactor (APWR) • Small modular reactor (SMR)

Utility-scale Storage • Pumped-hydro storage • Compressed air energy storage (CAES) Wind • Midcontinent Independent System Operator (MISO) • Southwest Power Pool (SPP) • In Valley • High voltage direct current (HVDC)

Coal fired • Integrated gas combined cycle (IGCC) • Supercritical pulverized coal 1x8 (SCPC1x8) • Supercritical pulverized coal 2x8 (SCPC2x8) • Integrated gas combined cycle with carbon capture and sequestration (IGCC CCS) • Supercritical pulverized coal 1x8 with carbon capture and sequestration (SCPC1x8 CCS) • Supercritical pulverized coal 2x8 with carbon capture and sequestration (SCPC2x8 CCS)

Solar • Utility-scale one-axis tracking photovoltaic • Utility-scale fixed-axis photovoltaic • Commercial-scale large photovoltaic • Commercial-scale small photovoltaic

Natural Gas fired • Simple cycle combustion turbine 3x (CT 3x) • Simple cycle combustion turbine 4x (CT 4x) • Combined cycle two on one (CC 2 by 1) • Combined cycle three on one (CC 3 by 1)

Biomass • New direct combustion • Repowering Energy Efficiency (EE) • Residential EE • Commercial EE • Industrial EE

Hydro • Hydro expansion project where spill permits • Hydro expansion project where space permits • Small-head or low-head (run of river) hydro project

Demand Response

Figure 5‑3: List of New Assets

See Chapter 4, Section 4.3.2, for a discussion of the different types of capacity ratings.

Nuclear There are three nuclear expansion options available to fill the expected capacity gap: a Pressurized Water Reactor (PWR), an Advanced Pressurized Water Reactor (APWR) and a Small Modular Reactor (SMR). The PWR option is based on completion of the Bellefonte brownfield site. The APWR and SMR options are not site specific.

TVA could increase the electrical output of the three Browns Ferry Nuclear units. This project could provide approximately 400 MWs of additional capacity and is termed an extended power uprate (EPU). Figure 5-4 provides an example of the characteristics for one of these projects. The book life is based on the remaining life of the plant.

Figure 5-4 shows some of operating characteristics used to model each option. Summer net dependable capacity, summer full load heat rate, unit availability and book life are explained above. The annual outage rate percentage includes forced and planned outages.

A nuclear PPA is also assumed to be available for model selection. PPAs are available for selection based on competitive information which cannot be disclosed. PPA

41



options are evaluated similar to build options with a few slight differences. One difference is when present value revenue requirements resulting from the expansion model selections are converted into cash flows, the build options have significant capital expenditures that match the construction spend schedule (noted

in section 5.1.2) versus the PPA options which have levelized cash flow payments based on the terms of the contract (similar to a mortgage). The other difference for PPAs is if the asset is located outside of the TVA transmission area then the necessary transmission wheeling charges are included.

PWR

APWR

SMR*

EPU 1

Summer Net Dependable Capacity (MW)

1,260

1,117

334

134

Summer Full Load Heat Rate (Btu/kWh)

9,853

9,715

10,046

9,558

Unit Availability (Yr)

2026

2026

2026

2018

Annual Outage Rate (%)

10%

10%

10%

10%

40

40

40

29

Unit Characteristics

Book Life (Yrs)

*The SMR option is based on a twin pack, the minimum viable configuration.

Figure 5‑4: Nuclear Expansion Options

42



BTUs of coal burned. The modeled CO2 emissions incur an emission penalty in the form of a dollar per ton of CO2 emitted.

Coal The 2015 IRP includes six coal expansion options, including two integrated gas combined cycle (IGCC) options and four supercritical pulverized coal (SCPC) options.

Two of the four SCPC options have one steam generator with a supercritical steam cycle. One of these options includes CCS technology; the other does not. The other two SCPC options have two steam generators with supercritical steam cycles. Again, one of these options includes CCS technology, and one does not.

IGCC technology converts coal into gas. One IGCC option has carbon capture and sequestration (CCS) and one does not. The CCS technology option is assumed to be commercially available starting in 2028 and has a 90 percent carbon dioxide (CO2) capture rate. Coal units typically have a CO2 emission rate of 205 pounds per million BTUs of coal burned so the CCS technology would reduce the CO2 rate to 20.5 pounds per million

Three options to continue to operate the Shawnee coal plant with the addition of more environmental controls (on various units) were available for model selection.

IGCC

IGCC* CCS

SCPC 1x8

SCPC 2x8

SCPC 1x8 CCS

SCPC 2x8 CCS


500

469

800

1,600

600

1,200


8,000

10,000

8,674

8,674

10,843

10,843


2022

2028

2025

2025

2028

2028


17%

18%

10%

10%

11%

11%

40

40

40

40

40

40


Book Life (Yrs)

*The CCS technology is assumed to have a 25% penalty on a 625 MW IGCC plant.

Figure 5‑5: Coal Expansion Options

43



All options are based on a generic location. The CO2 emission rate for a typical gas unit is 117 pounds of CO2 per million Btus of gas burned. The modeled gas units incur emission charges based on a dollar-per-ton emission penalty.

Natural Gas The IRP evaluation includes two simple cycle combustion turbine (CT) options and two combined cycle (CC) natural gas fueled options. The simple cycle CTs are available with either three or four turbines. The CC options have either two turbines and one steam generator (CC 2 by 1) or three turbines and one steam generator (CC 3 by 1). CC units have supplemental capacity termed duct-firing capacity that adds approximately 100 MW to the base capacity shown.

In addition, the IRP evaluation includes options for purchasing power from existing merchant gas plants, acquiring merchant gas plants, and options in which TVA would build additional gas-fueled units.

CT 3X

CT 4X

CC 2 by 1

CC 3 by 1


590

786

670

1,005


10,132

10,132

6,946

6,598

2018

2018

2019

2019


4%

4%

7%

7%

Book Life (Yrs)

30

30

30

30



Figure 5‑6: Gas Expansion Options

Petroleum Fuels TVA expects to phase out petroleum power purchases by 2028. There are no diesel fuels or other petroleum based resource options as a primary fuel source under consideration in the IRP because of emissions from these facilities.

44



Hydroelectric Two new hydro projects are included in the IRP evaluation, developed in collaboration with the TVRIX stakeholders. They include adding additional hydro turbines to existing dam facilities where there is space available with structural modifications. The other would add turbines at existing dam facilities where water that is now spilled could be used to power more turbines.

River system also are required for municipal and industrial uses, navigation, flood damage reduction, recreation, water quality and other purposes. For this reason, a fixed amount of monthly energy (provided by River Operations) is entered into the model for the conventional hydro stations. The model then uses the hydro energy to level the load shape served by other stations.

Both projects are similar to the larger TVA hydro system and are energy-limited units. Energy-limited units are resources that cannot be dispatched (in the model) based on price ($/MWh) as are traditional thermal generating resources, such as nuclear, coal and gas. Hydropower cannot be dispatched based on price alone because water releases in the Tennessee

Since hydro plants do not use fuel, a heat rate is not needed for modeling. Small- and low-head hydropower, called run of river, also is included as an IRP resource option. A hydro PPA was also included in the IRP evaluation.

Dam Spill Addition

Dam Space Addition

Run of River

40

30

25

2019

2018

2021

-

-

4%

40

40

40

Unit Characteristics Summer Net Dependable Capacity (MW) Unit Availability (Yr) Annual Outage Rate (%) Book Life (Yrs) Figure 5‑7: Hydro Expansion Options

45



underground cavern where it can be stored under pressure until electricity is required. The pressurized air is then heated and directed through a conventional generator to produce electricity.

Energy Storage The IRP evaluation includes a new hydroelectric pumped-storage unit as a resource option. The pumped-storage option would use three reversible turbine generators to either take electricity from the grid by pumping water into a higher altitude reservoir during periods of excess power or add electricity to the grid by using the pumped water to power a turbine as it falls from the upper to the lower reservoir.

Storage efficiency is included in modeling both these energy storage options because of the energy losses inherent to the energy conversion process and due to the loss of water or air during storage. The storage efficiency percentage for these energy storage options represents the efficiency of one cycle (i.e., pumping water, then releasing).

A compressed air energy storage (CAES) option also is included as an energy storage option. A CAES plant is similar to a pumped-storage plant but, instead of pumping water from a lower to an upper reservoir, a gas turbine is used to compress air often into an

TVA did not evaluate any electric battery storage options because of operational limitations.

Pump Storage

CAES


850

330


-

4,196

2023

2019


7%

10%

Storage Efficiency (%)

81%

70%

40

40



Book Life (Yrs) Figure 5‑8: Utility-Scale Storage Options

46



The HVDC option would require a third-party to permit and build a new transmission line so the unit availability is later than the other options. All unit availability dates were rounded to the next full year.

Wind Because TVA cannot take direct advantage of the tax credits and other investment incentives offered by the federal government to encourage wind power development, it has been more financially advantageous to acquire wind power resources through PPAs. This approach allows us to include wind as a resource option in the IRP. The purchase of wind resources as a PPA, whether produced in or imported to the TVA region, lowers the costs of these resources to TVA and its customers. TVA may evaluate the option of building wind facilities in the future if investment incentives and/ or future federal or state renewable mandates change.

Wind resources are energy- and capacity-limited resources. For this reason, we use an energy production profile to dispatch wind energy rather than price. The method used for wind resources is somewhat similar to hydro resources except that an hourly generation schedule (not a monthly amount) is pre-loaded into the capacity expansion model. We also apply a capacity credit since the total nameplate capacity of a wind turbine cannot be expected at the time of the system peak. To determine the capacity credit, we used historical data to estimate the typical wind power output at the time of the peak power demand on the TVA system. This resulted in a 14 percent capacity credit, meaning that 14 percent of nameplate capacity is expected to be available at the system peak. This reduced capacity is considered the summer net dependable capacity. Appendix B includes a more detailed discussion about the determination of the data assumptions for the modeling of the wind options included in this IRP.

Four wind options are included in the IRP evaluation, and the characteristics of these options were developed with input from the TVRIX stakeholders. The Midcontinent Independent System Operator (MISO), the Southwest Power Pool (SPP) and the In Valley options represent various wind resources in different regional transmission areas. The High Voltage Direct Current (HVDC) option would use a direct current (DC) bulk transmission system. The HVDC transmission system would reduce power losses that are typical of the more common alternating current (AC) transmission systems.

MISO

SPP

In valley

HVDC

Nameplate Capacity (MW)

200

200

120

250


28

28

17

35

2016

2016

2017

2020

-

-

-

-

20

20

20

20


Unit Availability (Yr) Annual Outage Rate Book Life (Yrs) Figure 5‑9: Wind Expansion Options

47



solar installations. The large and small scale commercial options represent solar installations at different price points and with different generating characteristics.

Solar Similar to new wind generation, because TVA cannot take advantage of the current investment incentives offered to promote solar power development, it is more financially advantageous to acquire solar power resources through PPAs. We may evaluate the option of building solar facilities in the future if investment incentives and/or federal or state renewable mandates change.

Like wind resources, solar resources are energylimited and therefore dispatched in the model using an hourly energy production profile to ensure that solar generation is not utilized by the model when the sun is not available. Solar resources also are similar to the capacity-limited wind resources where the availability of the unit at the time of the TVA system peak is less than the full nameplate capacity. We applied a 68 percent capacity credit for the utility tracking unit and a 50 percent capacity credit for the fixed asset options. The unit availability date was rounded to the first full year. More details about the assumptions used in the development of the unit characteristics for these solar options can be found in Appendix B.

Four solar options, developed with input from the TVRIX stakeholders, are included in the IRP evaluation at a minimum capacity block size of 25 MW nameplate capacity. All capacities are stated in alternating current (AC) terms. The utility tracking option is a single-axis tracker that allows the solar panels to follow the sun. The utility fixed option represents ground mounted fixed-axis/fixed-tilt

Utility tracking

Utility fixed

Commercial small

Commercial large

Nameplate Capacity (MW)

25

25

25

25


18

13

13

13

2015

2015

2015

2015

-

-

-

-

25

25

25

25


Unit Availability (Yr) Annual Outage Rate Book Life (Yrs) Figure 5‑10: Solar Expansion Options

48



the assumptions for hydro, wind, and solar, these options were also developed with input from the TVRIX stakeholders. Because biomass co-firing is considered a fuel switch opportunity, it was not included as a capacity expansion option.

Biomass Two new biomass options are included in the IRP evaluation: a new direct combustion biomass facility and a repower option, which is the conversion of existing coal-fired units to biomass-fired units. Like

Direct Combustion

Repower


115

75


13,500

12,243

2019

2015

Annual Outage Rate

5%

5%

Book Life (Yrs)

30

20



Figure 5‑11: Biomass Expansion Options

similar to those used for natural gas combustion turbines (CT). Demand response programs are operated much like CTs, or peaker units, and focus on reduction of peak demand. However, the terms of the demand response customer contracts are shorter than the expected book life of a CT unit.

Demand Response Demand response programs enable participating customers to reduce their power costs by allowing TVA to limit their power during peak demand times. Using a new innovative approach, these programs were modeled in the 2015 IRP based on unit characteristics

Demand Response Unit Characteristics Summer Net Dependable Capacity (MW)

1


10,132


2014

Annual Outage Rate

-

Book Life (Yrs)

5

Figure 5‑12: DR Expansion Options

49



EE “generating units” were developed to represent the residential (Res), commercial (Com) and industrial (Ind) sectors. Then each sector was divided into three tiers, representing three distinct price points, for a total of nine units. All of the tier 1 units are available beginning in 2014, but the first year tier 2 and 3 units will be available varies by sector. These units are energy limited, similar to hydro, wind and solar units, and use annual hourly production profiles.

Energy Efficiency The 2015 IRP reflects TVA’s increased focus on energy efficiency (EE). A new, innovative modeling approach was used in this IRP to evaluate EE as a supply-side resource, with characteristics and costs structured similarly to conventional generating resources or power plants. This allowed various EE “generating units” to be optimized against the other resource options. More details about this modeling approach can be found in Appendix D.

Res Tier 1

Res Tier 2

Res Tier 3

Com Tier 1

Com Tier 2

Com Tier 3

Ind Tier 1

Ind Tier 2

Ind Tier 3

10

10

10

10

10

10

10

10

10

-

-

-

-

-

-

-

-

-

2014

2022

2026

2014

2019

2022

2014

2018

2022

-

-

-

-

-

-

-

-

-

17

13

13

15

13

13

12

10

10

Unit Characteristics Nameplate Capacity (MW) Summer Full Load Heat Rate (Btu/kWh) Unit Availability (Yr) Annual Outage Rate Book Life (Yrs) Figure 5‑13: EE Expansion Options

50

Chapter 6

Resource Plan Development and Analysis Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

51

Re-evaluate

Recommend


Chapter 6: Resource Plan Development and Analysis

6 Resource Plan Development and Analysis

make (called planning strategies) in different sets of uncertain future conditions (called scenarios). The set of resource choices selected in any one future defines how we would provide power to our customers under those conditions; we call that set of resource choices a portfolio, and it is created by modeling a planning strategy in a particular scenario. These portfolios are then scored using some key factors (called metrics) that allow us to capture cost, risk, environmental footprint and other aspects that should be considered when deciding on the best target power supply mix.

This chapter describes the process TVA used to identify a target power supply mix that was based on the analysis done in the IRP. The process involves choosing the types of resources that we could use to meet the future power needs of our customers, recognizing that the future is uncertain and our choices need to give us flexibility to adapt. So the approach tests several options around resource choices we could

START

Develop Scenarios and Strategies

Develop Evaluation Scorecard

Resource Portfolio Optimization Modeling

Identify Preferred Target Power Supply Mixes

Incorporation of Public Input and Additional Modeling

Identify Recommended Target Power Supply Mix

END

Process for identifying the recommended target power supply mix

6.1 Development of Scenarios and Strategies

and a single scenario results in a resource portfolio7. A portfolio is a 20-year capacity expansion plan that is unique to that strategy and scenario combination.

TVA uses a scenario planning approach in integrated resource planning, a common approach in the utility industry. Scenario planning is useful for determining how various business decisions will perform in an uncertain future. The goal is to develop a least-cost strategy that is consistent with TVA’s legislatively mandated mission and also delivers our customers rate stability over a variety of future environments.

6.1.1 Development of Scenarios While most quantitative models used in long range planning focus on what is statistically likely based on history, market data and projected future patterns, TVA uses scenario analysis that allows for the possibility that the future could evolve along paths not suggested solely by historical trends.

Multiple strategies, which represent business decisions that TVA can control, are modeled against multiple scenarios, which represent uncertain futures that TVA cannot control. The intersection of a single strategy

The scenarios used in the IRP analysis were developed during the scoping phase of the study in 2013. The process used to develop these scenarios is described below. 7 Portfolios are also referred to as capacity expansion plans or resource portfolios

52



uncertainties. These uncertainties, shown in Figure 6-1, were used as building blocks to construct scenarios.

Identification of key uncertainties The first step in developing scenarios was to work with the individuals on the IRP Working Group to identify key

Description

Uncertainty TVA sales

The load to be served by TVA

Natural gas prices

The price of natural gas ($/MMBtu), including transportation

Wholesale electricity prices The hourly price of energy ($/MWh) at the TVA boundary (used as a proxy for TVA for market price of power) Coal prices

The price of coal ($/MMBtu), including transportation

Regulations

All regulatory and legislative actions, including applicable codes and standards, that impact the operation of electric utilities, excluding CO2 regulations

CO2 regulation/price

The cost of compliance with possible CO2 related regulation and/or the price of cap-and-trade legislation, represented as a $/Ton value

Distributed Generation

National trending of distributed generation resources and potential regional activity by customers or third-party developers (not TVA) See Appendix C for details on the method used to incorporate the effects of DG in the scenarios.

National Energy Efficiency (EE) adoption

An estimate of the willingness of customers nationally to adopt EE measures, recognizing the impacts of both technology affordability and electricity price

All aspects of the regional and national economy including general inflation, Economic outlook (national financing considerations, population growth, GDP and other economic and regional) drivers

Figure 6‑1: Key Uncertainties

53



• Placed sufficient stress on the resource selection process and provided a foundation for analyzing the robustness, flexibility and adaptability of each combination of supply- and demand-side options • Captured relevant key stakeholder interests.

Construction of Scenarios Scenarios were constructed using combinations of the key uncertainties shown in Figure 6-1 and then refined to ensure that each scenario: • Represented a plausible, meaningful future in which TVA could find itself within the 20-year study period • Was unique among the scenarios being considered for study

Figure 6-2 shows the key characteristics of the scenarios selected for the IRP analysis.

Scenarios

Key Characteristics

1 - Current Outlook

The outlook for the future which TVA is currently using for resource planning studies

2 – Stagnant Economy

Stagnant economy results in flat to negative growth, delaying the need for new generation

3 – Growth Economy

Rapid economic growth translates into higher than forecasted energy sales and resource expansion

4 – De-Carbonized Future

Increasing climate-driven effects create strong federal push to curb greenhouse gas emissions; new legislation caps and penalizes CO2 emissions from the utility industry and incentivizes non-emitting technologies

5 – Distributed Marketplace

Customers’ awareness of growing competitive energy markets and the rapid advance in energy technologies produce unexpected high penetration rates in distributed generation and energy efficiency. TVA assumes responsibility to serve the net customer load (no backup for any customerowned resources)

Figure 6‑2: Scenario Key Characteristics

54



The Current Outlook scenario projects growth of approximately 1.0 percent per year. Three scenarios – Stagnant Economy, De-Carbonized Future, and Distributed Marketplace – project lower load growth than the Current Outlook scenario, while the Growth Economy scenario models a modest growth scenario.

Determination of key scenario assumptions The final step in scenario development was to forecast key assumptions for each scenario. Figure 6-3 shows the forecasted assumptions for energy demand/load growth for each scenario.

Figure 6‑3: Energy Demand Assumptions

55



Figure 6-4 shows the forecasted assumptions for gas prices. Gas prices are similar for the Current Outlook, Stagnant Economy and Distributed Marketplace

scenarios, while both the Growth Economy and DeCarbonized Future scenarios assume a substantial increase in gas prices later this decade.

Figure 6‑4: Gas Price Assumptions

56



Figure 6-5 shows the forecasted assumptions for coal prices. Steadily increasing coal prices are forecasted for all scenarios. Starting in 2019, the De-Carbonized

Future scenario has the lowest price through the planning period.

Figure 6‑5: Coal Price Assumptions

57



Figure 6-6 shows the forecasted assumptions for CO2 prices. All scenarios forecast a more stringent regulatory future. The highest CO2 prices are seen in the De-carbonized Future scenario. The CO2 penalty in the Stagnant Economy scenario is the lowest and

does not start until 2029. The Current Outlook and Distributed Marketplace scenarios share the same CO2 price assumptions. Note that the CO2 cost curve for the Distributed Marketplace is the same as the assumptions used in the Current Outlook.

Figure 6‑6: CO2 Price Assumptions8 8 The cost curve for the Current Outlook and Distributed Marketplace are identical.

58



6.1.2 Development of Planning Strategies

lowest possible financial cost. TVA has developed strategies in both categories for this IRP. The process used to develop planning strategies is described below.

After the scenarios were developed, the next step in the IRP process was to design planning strategies. Scenarios and strategies are very different. Whereas scenarios describe plausible futures and include factors that TVA cannot control, strategies describe business decisions over which TVA has full control.

Identification of key strategy components The first step in developing planning Planning strategies was to identify the key strategies components, or attributes, to be represent included in each strategy. Ten distinct decisions and attributes were identified using input choices over from individuals on the IRP Working which TVA has Group and comments received control. during the public scoping period.

Generally, strategies fall into two categories: approaches that are intended to achieve a particular goal, but don’t restrict the energy resources that can be used to achieve that goal and approaches that constrain how resources are used. In IRP modeling terms, strategies that constrain resources are not fully optimized and may not produce plans that have the

59



Attributes

Description

Existing nuclear

Constraints related to TVA’s existing nuclear fleet, including Extended Power Uprates (EPUs)

Nuclear additions

Limitations on technologies and timing related to the addition of new nuclear capacity, including Watts Bar Unit 2, small modular reactors (SMRs), A/P 1000s and completion of TVA’s Bellefonte Nuclear Plant

Existing coal

Constraints related to TVA’s existing coal fleet, including the current schedule for idling coal units

New coal

Limitations on technology and timing on new coal-fired plants, including Carbon Capture & Sequestration (CCS) and Integrated Gasification Combined Cycle (IGCC) technologies

Gas additions

Limitations on technologies and timing related to the expansion options fueled by natural gas (CT, CC)

Energy Efficiency and Demand Response (EEDR)

Considers energy efficiency and demand response programs that are incentivized by TVA and/or local power companies, excluding impacts from naturally occurring efficiency/ conservation

Renewables (utility scale)

Limitations on technologies and timing of renewable resources, including options that could be pursued by TVA or in collaboration with local power companies

Purchased Power Agreements (PPAs)

Level of market reliance allowed in each strategy; no limitation on the type of energy source (conventional or renewable)

Distributed Generation/Distributed Energy Resources

Includes customer-driven resource options or third-party projects that are distributive in nature

Transmission

Type and level of transmission infrastructure required to support resource options in each strategy

Figure 6‑7: Key Planning Strategy Attributes

60



Development of Strategies Using Attributes TVA combined these 10 components to create five distinct planning strategies for the IRP analysis. Figure 6-8 lists the five strategies and their key characteristics.

Strategies

Key Characteristics

A – Traditional Utility Planning (Reference Plan)

Least cost optimization; EE/renewables selectable

B – Meet an Emission Target

Resources selected to create lower emitting portfolio based on an emission rate target or level using CO2 as the emissions metric

C – Focus on Long-Term, MarketSupplied Resources D – Maximize Energy Efficiency (EE) E – Maximize Renewables

Most new capacity needs met using longer-term PPA or other bilateral arrangements; TVA makes a minimal investment in owned assets Majority of capacity needs are met by setting an annual energy target for EE (priority resource to fill the energy gap); other resources selected to serve remaining need Enforce near-term and long-term renewable energy targets; meet targets with lowest cost combination of renewables; hydro is included as a renewable option along with biomass, wind and solar

Figure 6‑8: Planning Strategies Key Characteristics

Strategies A-C are strategies that achieve specific outcomes without setting any constraints or targets on the resource mix required to achieve the outcome. By contrast, strategies D and E are approaches that specify the way the resource mix will be constrained or how certain resource types must be prioritized in producing the resulting plan. Because strategies D and E are not fully optimized, they do not result in a plan that necessarily has the lowest financial cost. The latter two strategies were developed in collaboration with our stakeholders because these resource types

are getting more emphasis across the industry and we wanted to be able to answer questions about the benefits of increased EE and renewables in the mix. So these special purpose strategies will help inform TVA’s understanding about how the system would perform if priority were given to either EE or renewable resources to close the capacity gap. Definition of Strategies After defining each strategy’s key characteristics, specific descriptions were developed for its strategy attribute as shown in Figure 6-9.

61



STRATEGY ATTRIBUTES

Strategy A The ReferencePlan

Strategy B Meet An Emissions Target

Stategy C Focus on Long-Term, MarketSupplied Resources

Strategy D Maximize Energy Efficiency

Strategy E Maximize Renewables

Existing Nuclear

Operate existing units through end of period





Nuclear Additions

New nuclear is available for selection

New nuclear is available for selection

Only nuclear PPAs allowed.

No new nuclear

No new nuclear

Existing Coal

Based on current fleet strategy; 1) All coal units can be selected for retirement 2) SHF controls available for selection





New Coal

New coal allowed with CCS

New coal allowed with CCS

PPA is allowed

No additions

No additions

Gas Additions

Expansion option allowed


PPA is allowed



EEDR

EE and DR available for resource selection



EE required to meet all future energy needs first


Renewables (Utility Scale)

Expansion under current programs and new options available for selection




Aggressive renewable energy target enforced to promote growth in renewable resources first, through current programs or new options

New Energy Storage

Expansion options selectable


New energy storage not allowed



Hydro

Expension allowed; 1) PPA available 2) Capacity projects to existing assests available





Figure 6‑9: Strategy Descriptions

6.2 Resource Portfolio Optimization Modeling

Strategy attributes were used in the modeling in several different ways. For example, Strategy A has specific defined constraints such as new coal additions only with carbon capture and sequestration. Other components specified timing, such as allowing nuclear additions to be started after 2022 in Strategies A and B.

The generation of resource portfolios was a two-step process. First, an optimized portfolio, or capacity expansion plan, was generated, followed by a detailed financial analysis. This process was repeated for each strategy/scenario combination and for additional sensitivity runs.

62



6.2.1 Development of Optimized Capacity Expansion Plan

calculated detailed production costs of each plan including fuel and other variable operating costs. These detailed cost simulations provided total strategy costs and financial metrics that were used in the strategy assessment process.

TVA uses a capacity optimization model called System Optimizer.9 This model employs an optimization technique where an “objective function” (e.g., total resource plan cost) is minimized subject to a number of constraints.

This analysis was accomplished using a strategic planning software tool called MIDAS.10 MIDAS uses a chronological production costing algorithm with financial planning data to assess plan cost, system rate impacts and financial risk. It also uses a variant of Monte Carlo analysis,11 which is a sophisticated analytical technique that allows for risk analysis by varying important drivers in multiple runs to create a distribution of total costs rather than a single point estimate.

Energy resources were selected by adding or subtracting assets based on minimizing the present value of revenue requirements (PVRR). PVRR represents the cumulative present value of total revenue requirements for the study period based on an 8 percent discount rate. In other words, PVRR is the present day value of all future costs for the study period, discounted to reflect the time value of money and other factors such as investment risk.

The total cost for each resource plan (PVRR) was calculated taking into account additional considerations, including the cash flows associated with financing. The model generated multiple combinations of the key assumptions for each year of the study period and computed the costs of each combination. Capital costs for supply-side options were amortized for investment recovery using a real economic carrying cost method that accounted for unequal useful lives of generating assets.

In addition, the following constraints were applied in the optimization runs: • • • • • •

Balance of supply and demand Energy balance Reserve margin Generation and transmission operating limits Fuel purchase and utilization limits Environmental stewardship

In addition to computation of the total plan cost (PVRR) over the full 20-year study period, a 10-year system average cost metric was calculated. This metric provides an alternative view of the revenue requirements for the 2014-2023 timeframe expressed per MWh. It is not intended as a forecast of wholesale or retail rates over the study period. Rather, it was developed to gauge the potential rate impact associated with a given portfolio and provides an indication of relative rate pressure across the strategies being studied. A second system average cost metric covering the period 2024-2033 also was computed. Reviewing these two metrics in combination with PVRR and the financial risk measures provides a clearer picture of the cost/risk balance for each resource plan.

The System Optimizer model uses a simplified dispatch algorithm to compute production costs and a “representative hours” approach in which average generation and load values in each representative period within a week are scaled up appropriately to span all hours of the week and days of the months. Year-to-year changes in the resource mix are then evaluated and infeasible states are eliminated. The least cost path (based on lowest PVRR) from all possible states in the study period is used in the IRP as the optimized capacity expansion plan.

6.2.2 Evaluation of Detailed Financial Analysis Next, each capacity expansion plan was evaluated using an hourly production costing algorithm, which

10 MIDAS is also a Ventyx product. 11 Monte Carlo analysis is also referred to as stochastic analysis

9 System Optimizer is an industry standard software model developed by Ventyx.

63



6.2.3 Uncertainty (Risk) Analysis

Cost and risk metrics shown later in this report are computed based on the expected values produced from these stochastic iterations. The Midas tool allows TVA to explicitly consider uncertainty and risk exposure in the evaluation of the planning strategies. This analysis is based on applying probability distributions around the key variables used to frame the scenarios and define assumptions used in the strategies. The Monte Carlo analysis in MIDAS includes 13 key variables:

Stochastic analysis of production cost and financials bound the uncertainty and identify the risk exposure that is inherent in long-range power supply planning, because the fundamental forecasts used in those studies are inevitably wrong. Variability will result due to supply/demand disruptions, weather, market conditions, technology improvements, and economic cycles. A Monte Carlo simulation allows for a better understanding of the richness of possible futures, as well as their likelihoods so that plans can be made proactively as opposed to reactively. A stochastic model is used to estimate probability distributions of potential outcomes by allowing for simultaneous random-walking variation in many inputs over time.

• Commodity prices: natural gas, coal, oil, CO2 allowances, electricity price12 • Financial parameters: interest rates, capital costs, Operation and Maintenance (O&M) costs • Availability: hydro, fossil and nuclear • Load forecast uncertainty: demand and load-shape-year • Planning parameters: reserve margin target

At TVA, a representative Monte-Carlo distribution comprised of 72 stochastic iterations is developed for each of the scenario/strategy combinations to more fully assess the likely plan costs for each portfolio. A sample stochastic result is shown in Figure 6-10:

The fundamental (expected value) forecasts for these key variables differ across the five scenarios, and so the uncertainty ranges (stochastic envelope) are also different. So the evaluation of the uncertainty around the performance of the strategies considers both the variation across the scenarios (different plausible futures), as well as capturing the probability distribution around the expected forecasts represented by the stochastic envelope. As an example, Figure 6-11 shows these different uncertainty ranges around the TVA peak load forecast.

Example Stochastic Results

5th

Expected Value

95th

Figure 6‑10: Sample Stochastic Result

12 Stochastic electricity price was derived in MIDAS using stochastic variables as inputs

64



Figure 6‑11: Example Uncertainty Ranges The figure shows the range of variation in the expected forecast of peak demand across all five scenarios (represented by the gray shaded area); for orientation, the Current Outlook scenario’s fundamental forecast and its associated uncertainty range is shown in the black solid and dotted lines. The stochastic envelope, representing the uncertainty ranges from all five scenarios, is shown as the blue dotted line and bounds the uncertainty range evaluated in Midas. Each of the 13 key variables has a set of scenario ranges and stochastic envelopes that ensure a more dynamic assessment of the variability in the performance of each planning strategy.

the EE resource, we consider two primary sources of uncertainty: Design and Delivery Uncertainty. Design uncertainty exists for the following reasons: • Blocks are “proxies” for programs not yet developed some of which represent as-yet undeveloped technologies • Blocks are a blend of measures with different lifespans and each with a different underlying load shape Delivery uncertainty is driven by several factors: • The fact that TVA does not own the relationship with most end-use customers in the valley • Experience in other jurisdictions around nonperformance (realization rate) for both energy and demand • Uncertainty around the impact of future codes and standards on program design and deliveries (are EE program deliveries as certain in 2033 as they are in 2015?)

In addition to the uncertainty analysis based on the Monte Carlo modeling, in this IRP study we are including energy efficiency (EE) as a selectable resource. TVA made this decision to allow full portfolio optimization to clearly demonstrate value proposition and to allow flexible, nimble response to changing business environments. Uncertainty exists with all resource types and is modeled in different ways. For

65



A more complete discussion of the treatment of uncertainty for energy efficiency can be found in Appendix D. The uncertainty around the EE resource, combined with the Monte Carlo modeling of uncertainty, results in a robust evaluation of the planning strategies and allows TVA to more confidently apply metrics using a scorecard framework as a way to assess overall performance.

Strategic Imperatives Maintain low rates

Rates Asset Portfolio

6.3 Portfolio Analysis and Scorecard Development

Meet reliability expectations & provide a balanced portfolio

Modeling multiple strategies within multiple scenarios resulted in a large number of portfolios. So, initially, our portfolio analysis focused on common characteristics that strategies exhibited over multiple scenarios rather than on specific outcomes in individual portfolios. Strategies that behaved in a similar manner in most scenarios were considered to be “robust” – i.e., more flexible, less risky over the long-term and able to lessen the impacts of uncertainty. Conversely, strategies that behaved differently or poorly in most scenarios were considered more risky with a higher probability for future regret.

PEOPLE PERFORMANCE EXCELLENCE

Stewardship Be responsible stewards

Debt Live within our means

Figure 6‑12: Strategic Imperatives Optimizing TVA’s asset portfolio is the primary purpose of integrated resource planning, but other imperatives also shape the process: • As part of the financial analysis, a balance sheet and income statement are created for each portfolio to capture the rate revenues required to fund each resource plan. • A coverage ratio method is used to ensure that the overall debt limit is respected in each optimization run. • Stewardship obligations are considered in modeling of various compliance requirements, including portfolio optimization which factors in a carbon penalty and includes key environmental metrics in the assessment of each resource plan (air, water and solid waste impacts).

The first step in the portfolio evaluation process was to develop a scorecard to assess and compare the performance of planning strategies in each scenario. The process used to develop an evaluation scorecard is described below.

6.3.1 Selection of Metric Categories TVA’s mission and stakeholder concerns related to resource planning were key considerations in developing a set of metrics for use in evaluating the performance of the portfolios generated in the IRP. To achieve our overall mission of providing low cost, reliable power to the people of the Tennessee Valley, TVA focuses on four strategic imperatives: balancing rates and debt so that we maintain low power rates while living within our means; and recognizing the tradeoff between optimizing the value of our asset portfolio and being responsible stewards of the Tennessee Valley’s environment and natural resources.

As part of the public involvement process, stakeholders assigned priority to key concerns regarding the development of a long-range power supply plan, and priority concerns were used in identifying metric categories.

66



6.3.2 Development of Scoring and Reporting Metrics

Based on TVA’s strategic imperatives and feedback from stakeholders, five metrics categories were selected for use in evaluating the performance of planning strategies:

After establishing the metrics categories, the next step was to identify candidate metrics for each category. These metrics can be grouped into two broad categories:

• Cost, including both the long-range cost of the resource plan (present value of customer costs) as well as a look at a shorter term average system cost (an indicator of possible rate pressure) • Financial Risk, which measures the variation (uncertainty) around the cost of the resource plan by assessing a risk/benefit ratio and computing the likely amount of cost at risk using data from probability modeling • Environmental Stewardship, which captures multiple measures related to the environmental footprint of the resource plans such as air emissions and water or waste impacts • Valley Economics, which computes the macroeconomic effects of the resource plans by measuring the change in per capita income compared to a reference case • Flexibility, which measures how responsive the generation portfolio of each resource plan is by evaluating the type/quantity of resources and the extent to which this mix can easily follow load swings.

• Scoring metrics to be used in the scorecard to assess the performance of each strategy in different scenarios • Reporting metrics to be included in the IRP report as supplemental information for purposes of explanation and clarification. After considering the computational requirements and likely predictive value of multiple candidate metrics, as well as whether stakeholder groups would understand the purpose of each metric, TVA selected nine scoring metrics summarized in Figure 6-13.

67



Scoring Metric

Definition

20-year expected value PVRR

The total plan cost (capital and operating) expressed as the present value of revenue requirements over the 20-year study period (generated from the stochastic analysis, or the expected value of the probability distribution of plan costs)

Average system cost ($/MWh), Year 1-10

Average system cost for the first 10 years of the study, computed as the levelized annual average system cost (revenue requirements in each year divided by sales in that year)

Risk/benefit ratio

Area under the plan cost distribution curve between P(95)and expected value divided by the area between expected value and P(5)

Risk exposure

The point on the plan cost distribution below which the likely plan costs will fall 95% of the time based on stochastic analysis

CO2 annual average tons

The annual average tons of CO2 emitted over the study period

Water consumption

The annual average gallons of water consumed over the study period

Waste

The annual average quantity of coal ash, sludge and slag projected based on energy production in each portfolio

Flexibility

The annual system regulating capacity expressed as a percentage of peak load; measures the ability of the system to respond to load swings

% change in per capita income

The change in per capita personal income expressed as a change from a reference portfolio in each scenario

Figure 6‑13: Scoring Metrics

68



Figure 6-14 shows the formulas used to compute these scoring metrics. Category

Scoring Metric

Formula

PVRR ($Bn)

=

Present Value of Revenue Requirements over Planning Horizon

System Average Cost Years 1-10 ($/MWh)

=

NPV Rev Reqs (2014-2023) NPV Sales (2014-2023)

Risk/Benefit Ratio

=

95th (PVRR) – Expected (PVRR) Expected (PVRR) – 5th (PVRR)

Risk Exposure ($Bn)

=

95th Percentile (PVRR)

CO2 (MMTons)

=

Average Annual Tons of CO2 Emitted During Planning Period

Water Consumption (Million Gallons)

=

Average Annual Gallons of Water Consumed During Planning Period

Waste (MMTons)

=

Average Annual Tons of Coal Ash and Scrubber Residue During Planning Period

Flexibility

System Regulating Capability

=

Σ (Regulating Reserve + Demand Response + Quick Start) Peak Load

Valley Economics

Per Capita Income

=

Difference in the Change in Per Capita Personal Income Compared to Reference Case (for each scenario)

Cost

Risk

Environmental Stewardship

Figure 6‑14: Scoring Metric Formulas

69



In addition to the nine scoring metrics, seven reporting metrics were chosen: Reporting Metric

Definition

Average system cost ($/MWh), Year 11-20

Average system cost for the second 10 years of the study, computed as the levelized annual average system cost (revenue requirements in each year divided by sales in that year)

Cost uncertainty

The predicted variation in plan cost from the stochastic analysis, determined by using the difference between the tails of the distribution; the range in which plan costs will fall 90% of the time

Risk ratio

A measure of risk that the plan cost will exceed the expected value. This metric is developed by computing the ratio of the upper (higher cost) section of the cost distribution (between P(95) and the expected value) divided by the expected value

CO intensity

The CO emissions expressed as an emission intensity; computed by dividing 2 emissions by energy generated

Spent Nuclear Fuel Index

A measure of the quantity of spent nuclear fuel that is projected to be generated based on energy production in each portfolio

Flexibility

Two measures were selected in this category: the variable energy resource penetration, which measures the amount of variable or intermittent energy included in the plans; and a flexibility turn-down factor to measure the ability of the system to serve low load periods

Employment

The change in employment expressed relative to a baseline future

2

Figure 6‑15: Reporting Metrics

70



Figure 6-16 shows the formulas used to compute these scoring metrics. Category

Scoring Metric

Formula

Cost

System Average Cost Years 11-20 ($/MWh)

=

NPV Rev Reqs (2024-2033) NPV Sales (2024-2033)

Cost Uncertainty

=

95th (PVRR) – 5th (PVRR)

Risk Ratio

=

95th (PVRR) – Expected (PVRR) Expected (PVRR)

CO2 Intensity (Tons/GWh)

=

Tons CO2 (2014-2033) GWh Generated (2014-2033)

Risk


Expected Spent Fuel Generated During Planning Period

Spent Nuclear Fuel Index = (Tons) Variable Energy Resource Penetration

=

(Variable Resource Capacity) (2033) Peak Load (2033)

Flexibility Turn Down Factor

=

“Must run” + “Non-Dispatachable (Wind/Solar/Nuclear) (2033) Sales (2033)

Employment

=

Difference in the Change in Employment Compared to Reference Strategy

Flexibility

Valley Economics

Figure 6‑16: Reporting Metric Formulas

71



The scorecard metrics developed in collaboration with the IRP Working Group align with TVA’s mission as shown in Figure 6-17.

TVA Mission IRP Scorecard Metrics

Low-Cost Reliable Power

Economic Development


Technological Innovation

River Management

Present Value of Revenue Requirements System Avg. Cost Risk/Benefit Ration Risk Exposure CO2 Emissions Water Usage Waste Flexibility Impact to Per Capita Income

Figure 6‑17: Scorecard Alignment

6.3.3 Scorecard design

of the scenarios. Figure 6-18 shows the scorecard template, which includes nine columns (one for each of the scoring metrics, grouped by metric category) and five rows (one for each of the scenarios).

Once the scoring metrics were selected, the strategy scorecard could be designed. Using a format similar to the 2011 IRP, the scorecard summarizes the performance of an individual planning strategy in each

Figure 6‑18: Scorecard Template

72



The scorecard serves as a summary tabulation of the performance of the planning strategy in each scenario. To evaluate differences within a given scenario, all five scorecards should be reviewed. Interpretation of the performance of each strategy will be presented in Chapter 7.

6.4 Strategy Assessment Process Finally, scorecards were filled in based on an assessment of overall performance of each planning strategy in the five metric categories: cost, financial risk, stewardship, Valley economics and flexibility. Each metric category was assessed individually based on the simple average of the strategy’s performance in each scenario (assumes each scenario was equally likely), and graphics were developed to facilitate interpretation of trends and to identify preliminary observations. These observations will guide the development of an action plan for further case analysis. A cost/risk graphic was also prepared to enable an investigation of possible cost and risk trade-offs. The strategy assessment graphics, along with information about observations from the IRP study and the action plan, can be found in Chapter 8.

73

Chapter 7

Study Results Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

75

Re-evaluate

Recommend


Chapter 7: Study Results

7 Study Results

7.1 Analysis Results

This chapter describes the findings of the 2015 IRP. The results for 25 distinct portfolios are presented in this chapter along with the scorecard measures as discussed in Chapter 6.

7.1.1 Firm Requirements and Capacity Gap The key components of each scenario were translated into a forecast of firm requirements (demand plus reserves), which was used to identify the resulting capacity gap and need for power. This drove the selection of resources in the capacity planning model. Figure 7-1 illustrates the firm requirements forecasts for the five scenarios studied in the IRP.

Firm Requirements

MW, SND 45,000

40,000

35,000

30,000

2014

2016

2018

2020

2022

2024

Scenario 1: Current Outlook Scenario 3: Growth Economy Scenario 5: Distributed Marketplace

2026

2028

2030

2032

Scenario 2: Stagnant Economy Scenario 4: De-Carbonized Future

Figure 7‑1: Firm Requirements by Scenario Firm requirements were greatest in the Growth Economy scenario (highest load growth) and lowest in the Distributed Marketplace scenario (flat load growth until 2024). The remaining scenarios fell within this range and generally displayed smooth but unique growth trends, with the exception of the De-Carbonized Future scenario; the discontinuity exhibited in that scenario is the result of the abrupt application of an aggressive CO2 penalty.

The shape of the firm requirement curves influenced the type and timing of resource additions in the strategies. The timing of additional resources was a function of the existing system capacity and the impact of the attributes used to define each strategy. Figure 7-2 shows the range of the capacity gaps across the cases.13

13 Strategy assumptions are discussed in Section 6.1

76



Capacity Gap

MW, SND 15,000

10,000

5,000

0

(5,000)

2014

2016

2018

2020

2022

2024

Scenario 1: Current Outlook Scenario 3: Growth Economy Scenario 5: Distributed Marketplace

2026

2028

2030

2032

Scenario 2: Stagnant Economy Scenario 4: De-Carbonized Future

Figure 7‑2: Range of Capacity Gaps by Scenario

7.1.2 Expansion Plans

grouped together with the scenarios on the horizontal axis. For example, the first bar on the left of the chart is the incremental capacity results from the reference plan under the Current Outlook scenario. The incremental capacity additions are grouped by resource type (i.e., nuclear, hydro, coal, etc.).

The capacity expansion plans are presented below by strategy. Further information on the capacity expansion plans are presented in Appendix E – Expansion Plan Listing. Figure 7-3 presents the incremental capacity additions for all 25 cases by 2033. The ‘incremental’ capacity is the selected results that fill the capacity gap referenced above. The vertical axis is in summer net dependable (SND) megawatts, the capacity that can be applied to firm requirements. The results for each strategy are

The De-Carbonized Future and the distributed marketplace scenarios have the lowest demand forecasts and therefore have the least amount of incremental capacity. Conversely, the Growth Economy had the highest demand and therefore results in the most incremental capacity.

77



Incremental Capacity by 2033 MW, SND 14,000

Strategy A: Reference Plan

Strategy B: Meet an Emission Target

Strategy C: Market Supplied Resources

Strategy D: Maximize EE

Strategy E: Maximize Renewables

12,000 10,000 DR

8,000

EE

6,000

Gas CC

4,000

Renewables

Gas CT Coal

2,000

Hydro Nuclear

0

Figure 7-3: Incremental Capacity Additions for All 25 Cases

Capacity resource highlights are summarized below by resource type:

Economic Growth scenario (high growth scenario). This includes utility and commercial scale renewables but does not include small distributed renewable assets. The assumptions on distributed renewables were considered in the load demand projections for each scenario (see further discussion in Appendix C).

Nuclear: The extended power uprate (EPU) capacity expansion projects were selected in every case providing approximately 400 MW. No new nuclear was selected beyond the scheduled Watts Bar Unit 2.

Natural Gas: The addition of natural gas units vary more significantly than other resources and depend on the forecasted load in each scenario and the strategic focus. The maximum amount of additional CT capacity is approximately 5,000 MW in the high load world of the Growth Economy scenario. The lowest amount of additional CT capacity is about 200 MW in the Distributed Marketplace scenario. The incremental Gas CC capacity additions are similar across strategies A, B, D, and E with the Board-approved plans at Allen and Paradise and grow over time with the extension of existing contracts, the acquisition of contracted assets, or for new assets of about 800 MWs. In strategy C, a gas CC unit is replaced by a combination of controlled coal and more renewables and EEDR.

Hydro: Two of the smaller hydro capacity projects were selected in 24 of the cases. An additional hydro asset is typically selected across the Growth Economy and the de-carbonized future scenario, and also across the maximize renewable strategy cases. Coal: No new coal plants were selected. In a few cases, additional coal units were retired beyond those currently planned. Renewables: Figure 7-3 shows the non-hydro renewable assets (i.e., wind and solar) in summer net dependable megawatts which is the amount of firm capacity that can be expected at the system peak. The renewable additions range from ~1,000 MW to ~13,700 MW on a nameplate capacity basis. The lowest selection of renewable assets occurs in the Distributed Marketplace scenario (low growth scenario). The highest selection of renewables occurs in the

EE: The amount of energy efficiency added in strategies A, B, C, and E is fairly consistent averaging approximately 2,700 MW by 2033. The consistent

78



solar assets are selected in the mid-2020 timeframe and wind assets are selected in the late 2030 time period. However, in the De-Carbonized Future scenario, wind assets are selected as early as 2020. The natural gas assets increase over time, with the first addition occurring as early as 2020 in the Economic Growth scenario and as late as 2032 in the De-Carbonized Future scenario. The TVA Board-approved Paradise and Allen gas plants increase the gas portfolio by 2017 and 2019 respectively, then the percentage decreases over time as existing third-party contracts expire. In many scenarios existing contracts for gas combined cycle resources are renewed or the underlying asset is acquired. Energy efficiency increases in all scenarios decreasing the need for new gas resources. Demand response maintains a consistent portion of the capacity portfolio throughout the scenarios.

selection of energy efficiency is attributable to the low price compared to other assets and the energy contributions from the energy efficiency blocks. The one exception in this group is the Maximize Renewables strategy/Distributed Marketplace scenario case which has a lower selection of energy efficiency at 1,900 MW by 2033 due to the combination of low load assumptions and the strategic focus on renewables. The amount of EE in all of the strategy D/Maximize EE cases is also consistent at ~4,600 MW by 2033. DR: The incremental demand response averages out about 460 MW across all 25 cases with a range of almost 270 MW to 575 MW.

Summary by Strategy:

Figure 7-5 shows the energy portfolio which corresponds to the capacity charts in Figure 7-4. Nuclear energy increases over time due to the addition of Watts Bar Unit 2 and the extended power uprates. Hydro energy remains fairly constant. Coal generation decreases over the planning horizon as units are retired. The renewable generation remains fairly constant in the low demand scenarios (Stagnant Economy and Distributed Marketplace) and increases in the other three scenarios. Natural gas generation varies with load and strategic focus. Demand response, which produces low energy volumes, has been combined with the energy efficiency into one group termed EEDR. The incremental energy efficiency contributes 9% to 11% of the energy portfolio by 2033. Case 1A (the Current Outlook/Reference Plan case) results in 62% emission free energy by 2033.

Strategy A: The reference plan is TVA’s least-cost optimization plan and applies no special constraints or targets. Figure 7-4 presents the modeled capacity results for the reference plan. The capacity portfolios show the summer net dependable megawatts in 2033. The nuclear portfolio increases across all scenarios with the addition of Watts Bar Unit 2 and the extended power uprate projects. The hydro capacity increases slightly with the selection of projects that provide some additional capacity in all the cases. In the Growth Economy and the De-Carbonized Future scenarios, an additional hydro market asset is selected. The coal assets decrease in all scenarios by 2020 with announced retirements and decrease slightly more in the De-Carbonized Future and the Distributed Marketplace scenarios where additional coal units are retired. For most of the reference plan case results,

79


Chapter 7: Study Results Strategy A: The Reference Plan Capacity by 2033

Nuclear

Hydro

Coal

Renewables

Gas CT

Gas CC

EE

DR

Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbnized Future Scenario 5: Distributed Marketplace 0

10,000

20,000 30,000 MW, SND

40,000

50,000

*The nameplate capacity for the renewables category is as follows: Scenario 1: 5,050 MW, Scenario 2: 2,400 MW, Scenario 3: 7,500 MW, Scenario 4: 9,200 MW, Scenario 5: 2,200 MW.

Figure 7‑4: Capacity (Summer Net Dependable Megawatts) for Strategy A by Scenario

Strategy A: The Reference Plan Energy by 2033

Nuclear

Hydro

Coal

Renewables

Gas

EEDR

Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0

50

100 TWh

Figure 7‑5: Energy (Terawatt Hours) for Strategy A by Scenario

80

150

200

250



Strategy B: ‘Meet an Emission Target’ focuses on achieving a system-wide CO2 emission rate target in the least-cost manner. To set a target for the 20 year planning horizon ending in 2033, we used a glide slope that reduces TVA’s greenhouse gas emissions by 17 percent by 2020 and 80 percent by 2050 from a 2005 baseline. Strategy B adopts the 2033 data point on that glide slope of 557 pounds CO2 per MWh that translates to a 50 percent reduction in TVA’s system-wide CO2 emission rate from a 2005 baseline.

This strategy was not formulated to reflect EPA’s proposed Clean Power Plan (CPP) or rule. The proposed CCP was issued in June 2014 and its final form is uncertain. After EPA issues the final rule, States will have one to two years to decide how to implement it. The CCP is also expected to be litigated by others. TVA’s next update of its IRP will be able to take into account these developments. See section 7.1.3 for future discussion of the CCP.

The significant contributions from the selected energy efficiency and the renewable assets chosen in the reference plan result in reaching the CO2 emission target and therefore the two strategies are very similar. Figure 7-7 shows the energy portfolio for strategy B.

Figure 7-6 shows the capacity resources added by 2033 in strategy B across all five scenarios. The results from this strategy are very similar to the reference plan. The similarity of the case results was not anticipated during the development of the scenarios and strategies.

81


Chapter 7: Study Results Strategy B: Meet an Emission Target Capacity by 2033

Nuclear

Hydro

Coal

Renewables

Gas CT

Gas CC

EE

DR

Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace 0

10,000

20,000 30,000 MW, SND

40,000

50,000

* The nameplate capacity for the renewables category is as follows: Scenario 1: 5,925 MW, Scenario 2: 2,350 MW, Scenario 3: 8,300 MW, Scenario 4: 9,000 MW, Scenario 5: 2,200 MW.

Figure 7‑6: Capacity (Summer Net Dependable Megawatts) for Strategy B by Scenario

Strategy B: Meet an Emission Target Energy by 2033

Nuclear

Hydro

Coal

Renewables

Gas

EEDR


50

100 TWh

Figure 7‑7: Energy (Terawatt Hours) for Strategy B by Scenario

82

150

200



Strategy C: The Focus on Long-Term, Market Supplied Resources strategy is designed to constrain TVA capital spending in the least-cost manner. In this case, new self-build assets (i.e., TVA constructed) were restricted but improvements to existing owned assets and funds for energy efficiency and demand response programs were allowed, as well as power purchase agreements.

Figure 7-8 presents the total capacity portfolios for strategy C. The nuclear portfolio is similar to the reference plan strategy. The hydro assets increase above the reference plan by an additional 40 MW project in the Growth Economy and the De-Carbonized Future scenarios of strategy C. The coal portfolio increases slightly above the reference plan because maintaining existing coal resources is more favorable than procuring market supply. Third-party renewable assets increase above the reference plan since build options aren’t available for selection. Gas assets compete across the scenarios and selection depends on the scenario assumptions of load and commodity prices. The volumes on the gas assets selected in this strategy are slightly below the reference plan. Energy efficiency volumes remain similar across the scenarios as in the reference plan.

The original construct of strategy C in the Draft IRP allowed power purchase agreements of various lengths to be selected with contract terms ranging from as short as 10 years to as long as 20 years. However, current market assessments indicate a lack of market depth in the TVA territory and surrounding markets. To attract investment and to secure a project, however, is likely to require a longer-term commitment. Working with our stakeholder group, Strategy C was revised to include only 20 year contract terms and the results were updated. For further information see Appendix E.

The energy portfolio for this strategy is shown in Figure 7-9.

83



Strategy C: Market Supplied Resources Capacity by 2033

Nuclear

Hydro

Coal

Renewables

Gas CT

Gas CC

EE

DR


10,000

20,000 30,000 MW, SND

40,000

50,000

* The nameplate capacity for the renewables category is as follows: Scenario 1: 4,300 MW, Scenario 2: 2,950 MW, Scenario 3: 7,500 MW, Scenario 4: 9,100 MW, Scenario 5: 2,400 MW.

Figure 7‑8: Capacity (Summer Net Dependable Megawatts) for Strategy C by Scenario

Strategy C: Market Supplied Resources Energy by 2033

Nuclear

Hydro

Coal

Renewables

Gas

EEDR


50

100 TWh

Figure 7‑9: Energy (Terawatt Hours) for Strategy C by Scenario

84

150

200



Strategy D: The Maximize Energy Efficiency strategy requires that future energy needs be met first with EE in the least-cost manner.

to the reference plan. The hydro differs in scenario 4 (the decarbonized future) from the reference plan where additional EE replaces a market hydro asset. The coal portfolio varies in the Growth Economy and the De-Carbonized Future where the increase in EE results in additional coal unit retirements relative to the reference plan strategy. Renewables are reduced by about an average 400 MW SND in 2033 as compared to reference plan. Fewer natural gas units are selected relative to the reference plan given the increased deliveries from EE. Figure 7-12 shows the corresponding energy portfolios. Energy efficiency displaces an average of 11 terawatt hours of energy from coal, gas, and renewables as compared to the reference plan.

Figure 7-10 shows the energy efficiency capacity additions in strategy D across all five scenarios. The energy efficiency additions are categorized by enduse sector (i.e., industrial, commercial and residential). The amount of EE in strategy D increases above the reference plan strategy starting in 2024 and is approximately 1,900 MW and 11,200 GWh higher than the reference plan by 2033. Figure 7-11 shows the complete capacity portfolio for all of the strategy D cases. The nuclear assets are similar

Strategy D: Maximize EE MW, SND 5,000

Scenario 1: Current Outlook

Scenario 2: Stagnant Economy

Scenario 3: Growth Economy

Scenario 4: Decarbonized Future

Scenario 5: Distributed Marketplace

4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000

Industrial Commercial

500 0

Residental 2015 2020 2025 2030 2033

2015 2020 2025 2030 2033

2015 2020 2025 2030 2033

Figure 7‑10: Comparison of Energy Efficiency Resources in Strategy D

85

2015 2020 2025 2030 2033

2015 2020 2025 2030 2033



Strategy D: Maximize EE Capacity by 2033

Nuclear

Hydro

Coal

Renewables

Gas CT

Gas CC

EE

DR


10,000

20,000 30,000 MW, SND

40,000

50,000

*The nameplate capacity for the renewables category is as follows: Scenario 1: 2,825 MW, Scenario 2: 2,700 MW, Scenario 3: 7,200 MW, Scenario 4: 7,175 MW, Scenario 5: 1,025 MW.

Figure 7‑11: Capacity (Summer Net Dependable Megawatts) for Strategy D by Scenario

Strategy D: Maximize EE Energy by 2033

Nuclear

Hydro

Coal

Renewables

Gas

EEDR


50

100 TWh

Figure 7‑12: Energy (Terawatt Hours) for Strategy D by Scenario

86

150

200



Strategy E: The Maximize Renewables strategy enforces a renewable energy target of 20 percent by 2020 and 35 percent by 2040. The renewable energy target includes generation from new and existing hydro sources. The renewable energy strategy objective is met in the least-cost manner.

Figure 7-14 shows the capacity portfolios by 2033 for the strategy E cases. The nuclear assets are fairly similar to strategy A (the reference plan). Hydro increases above the reference plan strategy with the selection of a market asset in all 5 cases. The Maximize Renewables/Distributed Marketplace case with low loads, results in the most retired coal in the study. Renewables increase to more than 12 percent of the summer net dependable capacity portfolio by 2033 in all scenarios. The natural gas expansion is an average of 1,600 MW less than the reference plan across all scenarios.

Solar, wind and hydro resources were the renewable assets selected throughout the study. Figure 7-13 shows the new renewable additions by technology in five year increments for all five scenarios. The megawatts shown are the nameplate capacities. In Strategy E, hydro assets are added in every scenario. Wind is added by 2020 throughout the scenarios and almost doubles by 2033. Solar is selected in the nearterm at smaller amounts in the scenarios with some load growth. However, by 2025, the mix of renewables averages across the scenarios to be 6 percent hydro, 47 percent wind and 47 percent solar on a nameplate capacity basis.

MW, Nameplate 16,000

Figure 7-15 shows the corresponding energy portfolios. Renewable energy increases from an average of 17 terawatt hours in the reference plan strategies to 35 terawatt hours across strategy E.

Strategy E: Maximize Renewables Scenario 1: Current Outlook


Scenario 3: Growth Economy



14,000 12,000 10,000 8,000 6,000 4,000 Hydro

2,000

Solar 0

Wind 2015 2020 2025 2030 2033

2015 2020 2025 2030 2033

2015 2020 2025 2030 2033

Figure 7‑13: Comparison of Renewable Resources in Strategy E

87

2015 2020 2025 2030 2033

2015 2020 2025 2030 2033



Strategy E: Maximize Renewables Capacity by 2033

Nuclear

Hydro

Coal

Renewables

Gas CT

Gas CC

EE

DR


10,000

20,000 30,000 MW, SND

40,000

50,000

*The nameplate capacity for the renewables category is as follows: Scenario 1: 12,625 MW, Scenario 2: 12,125 MW, Scenario 3: 13,700 MW, Scenario 4: 11,450 MW, Scenario 5: 9,950 MW

Figure 7‑14: Capacity (Summer Net Dependable Megawatts) for Strategy E by Scenario

Strategy E: Maximize Renewables Energy by 2033

Nuclear

Hydro

Coal

Renewables

Gas

EEDR


50

100 TWh

Figure 7‑15: Energy (Terawatt Hours) for Strategy E by Scenario

88

150

200



7.1.3 The Clean Power Plan and the IRP

From a portfolio planning perspective, we think the TVA’s carbon emission rate is a better customerfocused planning metric for use in the IRP.

On June 2, 2014, the EPA issued proposed standards for carbon emissions from existing power plants, referred to as the “Proposed Clean Power Plan.” At the time of this publication going to print, EPA indicates the final rule will be released in August 2015. EPA received significant comments on the proposed rule, and the final rule is expected to change based on this input. The Proposed Clean Power Plan sets state-specific emission guidelines for carbon dioxide (CO2) emissions from power plants, targeting a 30 percent nationwide reduction in CO2 emissions from 2005 levels by 2030. Each state’s emission guideline is complex, and we refer you to EPA’s website14 for a detailed explanation.

The stringency of the Proposed Clean Power Plan’s state-by-state carbon emission numbers is established using four “building blocks” that together make up the “best system of emission reduction” or (BSER) for reducing carbon pollution. These are: (1) efficiency improvements of the coal fired plants themselves, (2) increased dispatch of Natural Gas Combined Cycles (NGCC) , (3) increased utilization of renewable energy, “at-risk nuclear,” and nuclear under construction and (4) increased demand-side energy efficiency. Each state’s emission guideline is calculated by applying these four building blocks to 2012 historical fossil emissions and generation. The EPA is proposing an “interim goal” that a state must meet on average over the 10-year period from 2020-2029 and a “final goal” that a state must meet at the end of that period in 2030 (and thereafter based on a three-year average). States must develop and submit plans to meet their goals and can comply individually or within a multi-state framework. As proposed, states would be required to submit their plans to the EPA by June 30, 2016. The final form of these standards is uncertain.

The Proposed Clean Power Plan’s emission guideline is in the form of a fossil energy, output-based, carbon dioxide (CO2) emissions rate for each state, which differs greatly from the system-wide carbon dioxide emission rates discussed in this IRP. A system-wide carbon dioxide emission rate, such as those reported in the IRP, is the amount of CO2 (as measured in pounds) that is emitted in the generation of a unit of electrical energy (a megawatt), expressed in lbs CO2 per MWh. As discussed above, the carbon emission rates in the proposed Clean Power Plan are primarily focused on fossil generation units only, and thus only include a portion of a utility’s total energy generation. In TVA’s case, the Clean Power Plan emissions rates do not include any of our hydro or most of our nuclear generation. Since collectively these non-CO2 emitting energy sources represent a large portion of the TVA system, the difference in targets between a systemwide CO2 rate and the Clean Power Plan targets for fossil-fueled resource are very large.

While the IRP models the amount of carbon contained in the delivered energy to our customers it does not model a potential compliance strategy for TVA with the Proposed Clean Power Plan. However, as a crude comparison, TVA has made a 30 percent reduction in CO2 emissions from a 2005 baseline, the stated objective of the regulation. One might assume that TVA would then have a low compliance hurdle with the CPP. However very much is unknown right now about how the final rule might change when promulgated this summer. For instance, in the proposed rule Tennessee’s emission guideline was made much more stringent than most all other states by considering Watts Bar Nuclear Unit 2 as an existing unit even though it is not yet operational. TVA has objected to this exclusion and we are awaiting the final regulation to see EPA’s final determination. Also, the final rule could change the timing of the regulation compliance period which would have a significant change in the stringency.

TVA believes that the use of the overall carbon rate is appropriate for the IRP because it simulates the amount of carbon that is in the delivered energy to our customers. This rate can be expressed as: TVA’s carbon emission rate = pounds of CO2 produced from power generation/total delivered energy MWh 14 http://www2.epa.gov/carbon-pollution-standards/clean-power-plan-proposed-rule

89



While we think that it is both inappropriate and premature to model the Clean Power Plan in the IRP, each of the five strategies significantly reduces carbon emissions on the TVA system through the planning period by retiring coal units and adding nuclear, renewables and energy efficiency. Considering these reductions and that compliance with the Clean Power Plan is further complicated by calling for state-bystate emission reductions although TVA’s integrated generation and transmission system encompasses parts of several states, it is inappropriate and premature to model the Clean Power Plan in this IRP. Regardless of the final form of the rule and which strategy TVA selects as a general planning direction from the IRP, we will be bringing the nation’s first new nuclear generation of the 21st Century online at Watts Bar 2 by 2016, retiring 13 units at two coal-fired power plants in Tennessee (ALF, JOF) and 13 units at two coal-fired power plants in Alabama (COF, WCF) by 2018, and replacing some of this coal generation with loweremitting natural gas, energy efficiency and renewables during the planning period. This will put TVA on a trajectory toward complying with the Clean Power Plan or regulatory requirements of another form in a carbonconstrained future.

and trends that can be informative to the preparation of the State Plans or assessing impacts from potential Federal Plans. The data, its underlying assumptions, and associated information must be used appropriately in venues that extend beyond this IRP.

7.2 Scorecard Results The fully populated scorecards for each of the five planning strategies are included in this section (see Chapter 6 for a discussion about the development of the scorecard template). Each strategy scorecard contains the metric values for that particular strategy in each of the five scenarios modeled in the IRP. The metric values are based on the combination of the portfolio optimization and uncertainty analysis work applied to each of the planning strategies under consideration. The scorecard for Strategy A is shown in Figure 7-16. The highest PVRR is the Growth Economy due to the large build-out to meet firm requirements. The highest system average cost is the De-Carbonized Future. The Growth Economy has the highest risk exposure driven by higher loads, and the Growth Economy has the highest CO2 releases, water consumption, and solid waste production. Note that the scorecard presents the system regulating capability snapshot in 2033 (more values for this metric are discussed in Chapter 8). Since the Valley economics metric uses Strategy A as the reference case in computing impacts, the change in per capita income is 0 percent for this strategy.

Neither Strategy B, nor any other strategy in this IRP study is intended to be a compliance strategy for the Clean Power Plan due to the aforementioned differences in carbon emission metrics. However, the findings and recommendations from this IRP will no doubt be useful in providing generalized observations

Figure 7‑16: Strategy A Scorecard

90



The scorecard for Strategy B is shown in Figure 7-17. These results are very similar to those shown for Strategy A, since the portfolios developed in that

strategy (and in particular the contribution from energy efficiency and renewables) generally achieve the overall system emission target designed for Strategy B.

Figure 7‑17: Strategy B Scorecard

The scorecard results for Strategy C are shown in Figure 7-18. PVRR costs are slightly higher than Strategy A reflecting the increased cost associated with the third-party PPA capacity additions in this strategy. Compared to Strategy A, system average cost metrics are slightly better (actually lowest overall), and risk exposure is just slightly higher. This strategy has the

highest environmental impacts for CO2 and water consumption, primarily caused by the higher fraction of fossil-fueled generation in this case. Flexibility scores are lower compared to the results for Strategy A due to the lower percentage of regulating capacity added in this plan (the portfolio has a higher contribution from long-term PPAs and retained coal capacity).

Figure 7‑18: Strategy C Scorecard

91



The Strategy D scorecard is shown in Figure 7-19. PVRR cost rankings across the scenarios are similar to strategy A but total costs are generally higher. System average costs are similar in the first 10 years. The Risk/ Benefit Ratio and Risk Exposure are higher for this

strategy, due to the requirement that resource needs be met first with energy efficiency, thereby restricting portfolio composition. However, this strategy has better performance in environmental metrics.

Figure 7‑19: Strategy D Scorecard

Strategy E metric values are shown in Figure 7-20. PVRR costs are higher than respective strategy A costs for all cases, the result of aggressive renewable resource targets. Correspondingly, system average costs are higher in all strategy E cases. The Risk/ Benefit Ratio and Risk Exposure are higher with more

renewables, which indicates that the enforced targets may be too high relative to the benefits derived from adding renewable resources to the portfolio. This strategy has the best performance in all environmental metrics, driven by the higher concentration of renewable resources in the cases.

Figure 7‑20: Strategy E Scorecard

92



7.3 Scoring Metric Comparisons Figure 7-21 shows a comparison of how each strategy scored across all scenarios by metric

Scenario

PVRR ($ billion)

System Average Cost 2014-2023 ($/MWh)

Risk Exposure ($ billion)

Risk Exposure ($ billion)

CO2 Emissions (million tons/year)

Water Consumption (million gallons/year)

Alternative Strategy

1 Current Outlook

2 Stagnant Economy

3 Growth Economy

4 De-Carbonized Future

5 Distributed Marketplace

Average

A

132.7

125.9

139.5

131.7

120.4

130.0

B

132.7

126.0

139.5

131.7

120.4

130.1

C

133.4

126.5

140.8

131.9

121.1

130.7

D

134.4

127.9

141.3

133.6

122.8

132.0

E

136.2

129.4

140.8

132.8

123.5

132.5

A

76.7

76.0

77.7

81.0

77.3

77.7

B

76.7

76.0

77.7

80.8

77.3

77.7

C

76.4

75.7

77.8

80.6

76.8

77.5

D

76.9

75.9

77.5

81.1

77.3

77.7

E

78.4

77.3

78.5

81.3

78.5

78.8

A

0.92

0.95

0.91

1.00

0.99

0.95

B

0.92

0.95

0.92

0.99

0.99

0.95

C

0.90

0.95

0.91

0.98

0.99

0.95

D

0.94

0.98

0.92

1.03

1.00

0.98

E

1.03

1.04

1.03

1.01

1.05

1.03

A

140.4

132.8

147.5

140.3

127.1

137.6

B

140.4

133.0

147.6

140.3

127.1

137.7

C

141.2

134.0

149.3

140.7

128.3

138.7

D

142.4

135.3

149.7

142.7

130.0

140.0

E

145.1

137.4

149.8

141.7

130.9

141.0

A

57.0

51.8

59.7

44.2

44.2

51.4

B

57.0

51.8

59.7

44.3

44.2

51.4

C

58.2

52.9

59.1

44.2

45.2

51.9

D

56.2

50.7

57.6

41.8

43.5

50.0

E

52.2

45.6

54.2

41.6

39.9

46.7

A

72,952

73,429

79,489

65,890

67,536

71,859

B

72,988

73,438

79,508

66,001

67,533

71,894

C

74,504

75,042

79,296

66,227

68,840

72,782

D

72,657

72,827

77,482

65,696

67,020

71,136

E

69,019

68,389

74,733

65,450

63,799

68,278

93


Chapter 7: Name of Chapter

Scenario

Waste (million tons/year)

System Regulating Capability (2033)


1 Current Outlook

2 Stagnant Economy

3 Growth Economy

4 De-Carbonized Future

5 Distributed Marketplace

Average

A

3.46

3.49

3.72

3.08

3.21

3.39

B

3.46

3.50

3.71

3.10

3.21

3.39

C

3.41

3.46

3.69

3.09

3.24

3.38

D

3.44

3.44

3.73

2.75

3.17

3.31

E

3.16

3.13

3.50

2.75

2.93

3.10

A

28.7%

28.0%

27.1%

18.9%

22.3%

25.0%

B

29.9%

27.9%

26.2%

19.7%

22.3%

25.2%

C

28.5%

26.0%

28.6%

20.4%

18.2%

24.4%

D

27.7%

22.3%

26.4%

20.3%

25.0%

24.3%

E

20.9%

20.4%

23.5%

18.8%

16.0%

19.9%

B

0.00%

0.01%

-0.01%

0.00%

0.00%

0.00%

C

0.00%

0.01%

0.03%

0.01%

0.00%

0.01%

D

0.02%

0.02%

0.02%

0.02%

0.02%

0.02%

E

-0.01%

0.00%

0.00%

0.00%

-0.01%

-0.01%

A Percent Difference in Per Capita Income (Relative to Strategy A)

Figure 7‑21: Scoring Metrics by Strategy & Scenario

7.4 Preliminary Observations

• Higher levels of energy efficiency and renewable resources are indicated in many cases over the 20 year study period • Changing environmental standards for CO2 will drive retire/control decisions on some coal-fired generation in the mid-2020s • Solar resources begin appearing in the resource plans in the mid 2020s; wind resources appear in the late 2020s in some scenarios, and generally the HVDC wind option is not selected until early 2030s

Based on the results of the modeling to date, TVA made some observations about the case results: • There is a need for new capacity in every scenario being modeled, even in the lower load futures • There are no immediate needs for baseload resources beyond the completion of Watts Bar Unit 2 and the Browns Ferry extended power uprates • Most of the variation in expansion plans is around natural gas and renewables and most of the resource plans show a tradeoff between EE and gas resources

These observations are further explored in the assessments presented in Chapter 8.

94

Chapter 8

Strategy Assessments Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

95

Re-evaluate

Recommend


Chapter 8: Strategy Assessments

8 Strategy Assessments

• Expected Value PVRR, 20 Year – the total plan cost (capital and operating) expressed as the present value of revenue requirements (PVRR) over the 20year study period. These metrics allowed us to compare the cost and financial risks associated with different planning strategies from both a short-term (10-year) and a longterm (20-year) perspective. (See Chapter 6, section 6.2.2, for more information on scoring metrics, including the formulas used to compute them.)

This chapter explains the strategy assessments and summarizes the results. Areas where additional study may be needed and next steps in the IRP process are also discussed.

8.1 Strategy Assessments To assess the performance of the five planning strategies (explained in Chapter 6 and shown below), we used scorecard data to conduct four assessments: • • • •

Figure 8-1 shows the results for the 10-year system average cost metric. The blue bar represents the system average cost values for the first 10 years in the study period (2014-2023), and the red bar represents the second 10-year period (2024-2033).

Cost and risk Environmental stewardship Flexibility Valley economics

We calculated the overall value of each strategy by averaging its performance over every scenario, since all of them are presumed to be equally likely.

8.1.1 Cost and Risk Assessment The cost and risk assessment was aimed at gaining a better understanding of the relative performance of different strategies in terms of total plan costs and financial risk. The cost assessment was based on two scorecard metrics: Planning Strategies • System Strategy A: Traditional Utility Planning Average Cost ($MWh), Year Strategy B: Meet an Emissions Target 1-10 – the Strategy C: Focus on Long-Term, average system Market Supplied Resources cost for the first 10 years of the Strategy D: Maximize Energy Efficiency study, computed Strategy E: Maximize Renewables as the levelized annual system average cost (i.e., revenue requirements in each year divided by sales in that year)

Figure 8‑1: System Average Cost During the first 10-year period, the system average cost is essentially the same across all five strategies. However, in the second 10-year period, there is some variation, with Strategy D exhibiting the highest system average cost. This is likely the result of increased costs for energy efficiency programs combined with a resultant reduction in energy sales. These factors combine to shrink sales and put upward pressure on the system average cost. Figure 8-2 shows the results for the 20-year present value revenue requirement (PVRR) metric. The chart

96



shows the range of plan costs as well as the expected value for each strategy across all the scenarios. The lower end of each bar is the best case (lowest cost) outcome from the uncertainty analysis; the upper end is the worst case (highest cost) outcome; the expected value is the point of transition between the two colored sections of each bar. Strategies A, B and C have roughly the same average PVRR results across all scenarios and the lowest set of total plan costs measured in terms of the 20-year PVRR. Strategies A

and B are virtually identical due to the selection of EE in Strategy A which enables that case to essentially achieve the emission target set in Strategy B, while Strategy C is slightly more expensive. Strategies D and E, which constrain the selection of resource types used in the plan, are projected to have a PVRR that is about $2 billion higher over the 20-year planning period, while the range of possible outcomes for all five strategies is fairly consistent as shown by the height of the bars.

Figure 8‑2: Total Plan Cost (PVRR)

Two additional metrics were used to assess the risk of each strategy:

area between Expected Value and P(5) (when costs are less than the expected value)

Risk/Benefit Ratio – the area under the plan cost distribution curve between P(95) and Expected Value (when costs exceed the expected value) divided by the

Risk Exposure – the point on the plan cost distribution below which the likely plan costs will fall 95 percent of the time (this is also the worst-case outcome).

97



Figure 8‑3: Risk/Benefit Ratio

Figure 8-3 shows the risk/benefit ratios for the five planning strategies. In this metric, lower values indicate better performance where the benefits outweigh the risks. Risk/benefit scores less than 1.0 indicate that costs are more likely to be less than the expected value.

value, caused in part by the aggressive renewable targets established in this case. We investigated key assumptions in Strategy E in an effort to better understand this result, and those results are discussed in Section 8.3.

Strategies A-C have very similar scores, with Strategy C scoring just slightly lower (better performance) than Strategies A/B. Strategy E appears to be the most risky from a financial perspective. It is the only strategy with a ratio greater than 1.0, indicating that plan costs in this strategy are more likely to exceed the expected

Figure 8-4 shows TVA’s risk exposure under the five strategies. This metric measures the worst-case outcome as represented by the P(95) value of the PVRR distribution and is useful in determining which strategies present the higher financial risks.

98



Figure 8‑4: Risk Exposure

Strategies A and B have essentially the same risk exposure; the risk exposure for Strategy C is higher by about $1 billion, reflecting the potential upside cost associated with the long-term power agreements in that case. Strategies D and E have distinctly higher exposure values – as much as $3.5 billion higher. This indicates that strategies where resource selection is constrained, such as aggressive targets imposed for energy efficiency (Strategy D) and renewables (Strategy E) carry higher financial risks than the other three strategies. In both of these strategies, the required resource contributions (EE or renewables) tend to limit the flexibility to optimize a portfolio in the uncertainty

analysis, leading to these higher financial risk scores. Strategy E has the highest risk exposure and is also the only strategy with a risk/benefit ratio greater than 1.0. This indicates that this strategy may be the most risky financially of those evaluated in the IRP. This result is driven by the very aggressive targets for renewable resources that are imposed in the strategy. Another way to assess cost and financial risk is to combine the cost and risk scores so a trade-off analysis can be performed. Figure 8-5 shows cost/risk tradeoffs based on total plan cost and system average cost.

99



Figure 8‑5: Cost/Risk Trade-Offs

100



These charts also reinforce the cost and risk assessment results discussed about Strategies D and E having somewhat higher plan costs and exhibiting higher financial risks, with Strategy E showing the highest cost and risk outcome. There is also a tradeoff between Strategies A/B and C in the upper chart in Figure 8-5, which indicates that a somewhat improved (lower) risk/benefit ratio can be achieved for these strategies but at a slightly higher plan cost.

• CO2 Average Tons – the annual average tons of CO2 emitted over the study period • Water consumption – the annual average gallons of water consumed over the study period • Waste – the annual average quantity of coal ash, sludge and slag based on energy production in each portfolio. Figure 8-6 shows the average environmental impact for each strategy for each of these three metrics. The graphic presents the impacts on a relative basis, normalized to the highest impact for each metric. More information about the development of these metrics can be found in Appendix F.

8.1.2 Environmental Stewardship As discussed in Chapter 6, strategy scorecards include three measures for environmental stewardship performance:

Figure 8‑6: Environmental Impacts

Strategies A, B and C have almost the same environmental impacts across all three metrics, with Strategy C having a slightly higher impact. Strategy D shows somewhat lower environmental impacts for

all three metrics, with Strategy E showing the lowest impacts. The air and waste impacts in Strategy E are significantly lower than the other strategies due to the emphasis on renewable resources.

101



8.1.3 Flexibility

This is the first time TVA has used annual system regulating capability as a metric to assess the performance of a resource portfolio, and further work is planned after the completion of this IRP to determine what the minimum or optimum flexibility score should be for the TVA system.

Annual system regulating capability, expressed as a percentage of peak load, was used to measure the flexibility of the five planning strategies. TVA considers flexibility – the ability of the system to respond to load swings – as a key future consideration for long-range resource planning.

Figure 8-7 shows flexibility scores for each strategy at three points within the study window: 2014, 2024 and 2033 (higher is better).

This is especially true as the resource mix shifts from traditional, fully dispatchable central station units toward more diverse and dispersed generating assets.

Figure 8‑7: System Regulating Capability

Strategy D has a higher flexibility score during the first ten years of the study period due to lower system load. However, during the second decade, the quick response units added in Strategies A and B result in similar levels of regulating capability. By the end of the study period, Strategies C and D have slightly lower flexibility scores, likely the result of fewer quick-start

capacity additions due to the higher commitment to long-term PPAs or energy efficiency resources in those strategies. The results for Strategy E are significantly different because this strategy has a higher percentage of non-dispatchable renewable resources and thus a reduced ability to respond to unexpected load swings.

102



8.1.4 Valley Economics

income level established by Strategy A in each scenario. More details about how TVA has computed this macroeconomic impact can be found in Appendix G. The results are shown in Figure 8-8.

The impact of different planning strategies on the Valley economy was assessed based on the percent change in per capita income, measured from the reference

Figure 8‑8: Valley Economics

Strategy D consistently outperformed the reference income level across all scenarios. This is likely due to the retention of more investment in the Valley under this strategy driven by the commitment to energy efficiency, which results in increased investment in the Valley

relative to other resource options. However, the overall variation in per capita income estimates is very small across the strategies. This indicates that the Valley Economics metric is unlikely to be a key consideration when selecting a preferred target power supply mix.

103



8.1.5 Summary of Initial Observations The overall performance of the five planning strategies is summarized by metric category in Table 8-1 and by strategy in Table 8-2.

Metric Category

Assessment Observations

Cost

On the basis of average system costs, all five strategies are very similar over the first 10-years of the study period. Total plan costs over the 20-year study period are also similar, with the strategies D and E that constrain the selection of resource types in the plan, more expensive.

Financial Risk

Risk scores are lower for the strategies that emphasize significant investment in any one particular technology. For example, we see higher risk in portfolios that focus on higher levels of energy efficiency or renewables. (Note: sensitivity cases found similar risk profiles for portfolios that concentrated on nuclear technologies.)


All strategies show significant improvement in air (CO2), water and waste categories compared to the performance of the current resource portfolio, with the Maximize Renewables strategy having the lowest environmental impact.

Flexibility

All strategies appear similar, but the ability of the system to respond to load uncertainty is most limited in the Maximize Renewables strategy. The flexibility score for the Maximize Energy Efficiency strategy is likely a result of reduced loads.

Valley Economics

All strategies seem to have comparable impact on the Valley economy as measured by per capita income. The Maximize Energy Efficiency strategy appears to have a slightly stronger economic impact due to a higher percentage of investments remaining in the Valley.

Table 8‑1: Summary of Observations by Metric Category

104



Strategy

Assessment Observations

Strategy A: Reference Plan

• Relatively low PVRR and System Average Cost during the first 10 years of the study period • Lowest System Average Cost in the second 10 years of the study period • Low financial risk (risk/benefit ratio less than one; second lowest risk exposure) • Higher environmental impact compared to Strategies D and E • Demonstrates flexibility

Strategy B: Meet an Emission Target

• Results are nearly identical to Strategy A

Strategy C: Rely on Long-Term, Market-Based Resources

• Slightly higher PVRR, relatively low system average cost, and moderate financial risk • Higher environmental impact than other strategies • Somewhat lower system regulating capability than Strategies A or B

Strategy D: Maximize Energy Efficiency (EE)

• Higher PVRR than Strategies A, B or C • Relatively similar system average cost to other strategies during the first decade, but high system average cost during the second decade due to increasing levels of EE and lower power sales • Comparable to Strategy C on flexibility performance due to reduced sales • Low environmental impact, second only to Strategy E

Strategy E: Maximize Renewables

• • • •

Highest PVRR in all scenarios due to enforcement of renewable energy targets Highest risk/benefit ratio of any strategy (greater than 1.0) Lower flexibility performance than other strategies Lowest environmental impact

Table 8‑2: Summary of Observations by Strategy

105



8.2 Reporting Metrics Comparisons

performance of the individual planning strategies in each of the results. Figure 8-9 shows a comparison of how each strategy scored across all scenarios by reporting metric.

As further described in Chapter 6, in addition to scoring metrics, reporting metrics were selected to provide further explanation and clarification in interpreting the

System Average Cost 2024-2033 ($/MWh)

Cost Uncertainty ($Bn)

Risk Ratio


Spent Nuclear Fuel (Tons/Year)

Variable Resource Penetration 2033

Scenario


1

2

3

4

5

Average

A

98.7

94.8

100.4

103.0

98.7

99.1

B

98.6

95.1

100.4

103.4

98.7

99.3

C

100.5

96.7

104.2

104.6

101.4

101.5

D

104.5

102.4

106.8

110.0

108.3

106.4

E

102.0

99.1

100.8

104.6

101.6

101.6

A

16,014

14,331

16,810

17,277

13,435

15,573

B

16,051

14,295

16,884

17,241

13,422

15,579

C

16,538

15,318

17,910

17,940

14,628

16,467

D

16,477

15,008

17,420

17,919

14,296

16,224

E

17,527

15,677

17,751

17,664

14,589

16,642

A

0.058

0.055

0.057

0.065

0.056

0.058

B

0.058

0.055

0.058

0.065

0.055

0.058

C

0.059

0.059

0.061

0.067

0.060

0.061

D

0.059

0.058

0.059

0.068

0.058

0.061

E

0.065

0.062

0.064

0.067

0.061

0.064

A

350.0

330.0

352.9

291.3

306.9

326.2

B

350.3

330.1

353.0

292.2

306.9

326.5

C

357.5

337.4

352.2

291.4

314.4

330.6

D

351.4

329.6

345.6

279.9

308.5

323.0

E

320.4

290.7

319.8

273.5

275.1

295.9

A

149.05

149.05

149.05

149.05

149.05

149.05

B

149.05

149.05

149.05

149.05

149.05

149.05

C

149.05

149.05

149.05

149.05

149.05

149.05

D

149.05

149.05

149.05

149.05

149.05

149.05

E

149.05

149.05

149.05

149.05

149.05

149.05

A

24.9%

17.8%

31.2%

40.7%

19.2%

26.8%

B

24.7%

17.6%

33.5%

39.6%

19.2%

26.9%

C

22.9%

20.0%

31.8%

40.9%

20.0%

27.1%

D

19.7%

19.6%

32.2%

36.5%

16.9%

25.0%

E

48.1%

48.1%

48.1%

48.7%

45.0%

47.6%

106



Flexibility Turndown Factor 2033

Scenario


1

2

3

4

5

Average

A

49.1%

45.2%

52.5%

65.0%

52.9%

53.0%

B

48.9%

45.2%

54.1%

64.6%

52.9%

53.2%

C

46.1%

45.9%

52.8%

64.9%

53.4%

52.6%

D

46.7%

48.8%

55.5%

65.0%

56.4%

54.5%

E

62.2%

63.3%

60.6%

66.4%

68.0%

64.1%

B

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

A NonFarm Employment: % Changes from Reference Plan (A)

C

0.0%

0.0%

0.1%

0.0%

0.0%

0.0%

D

0.1%

0.1%

0.1%

0.1%

0.1%

0.1%

E

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Figure 8‑9: Reporting Metrics by Strategy & Scenario

8.3 Sensitivity Analysis

2. Energy Efficiency and Demand Response sensitivities that tested the impact of key energy efficiency cost and performance assumptions and assessed the impact of energy efficiency and demand response on the overall portfolio. 3. Renewables sensitivities that evaluated the impact of key pricing and performance assumptions around renewable technologies. 4. Resource sensitivities that tested the impact to the case results of adding other resources not selected in the initial runs. 5. Key Driver sensitivities that analyzed the impact to the case results if a specific combination of assumptions was imposed on the optimization model, rather than using the correlated scenario assumptions developed for the study. An example would be forcing in a high gas price forecast.

During the course of developing the draft IRP, TVA identified questions and findings that warranted further evaluation prior to finalizing the study. In addition, we received helpful stakeholder feedback from the IRP Working Group (IRPWG), the Regional Energy Resource Council (RERC), and through our public meetings and formal comment period that helped identify key areas that merited further analysis. To address these questions and comments we performed detailed sensitivity analyses which were reviewed with the IRPWG and the RERC in April of 2015. The sensitivity cases generally fell into five primary categories: 1. Nuclear sensitivities that tested the impact to the case results if different nuclear options not selected in the initial case runs were forced into the portfolio.

107



The sensitivity cases are listed below in Figure 8-10.

Key findings are summarized below by sensitivity case category:

Nuclear Sensitivities • Pressurized water reactor or Bellefonte Unit 1 and Unit 2 • Advanced pressurized water reactor or AP 1000 • Small modular reactor

Nuclear Sensitivity Results: This set of cases examined the impact of forcing in Bellefonte unit 1 in 2026, Bellefonte 1 and 2 in 2028, an AP 1000 in 2028, and a Small Modular Reactor into the resource plan in 2028. These resources were added in the year specified and the portfolio was re-optimized within the framework of Scenario 1 and Strategy A. Conclusions are as follows:

Energy Efficiency and Demand Response Sensitivities • No Energy Efficiency Planning Factor Adjustment • Faster Energy Efficiency Ramp Rate • No Demand Response • No Energy Efficiency or Demand Response

• New nuclear additions result in higher overall system costs than the reference plan but would deliver value beyond the study window. Cost-sharing mechanisms that could be made available for Small Modular Reactors (SMRs) have not been included but, if available, would render those options more financially attractive. • Short-term system average costs are higher with nuclear builds, but long-term average costs are similar to non-nuclear cases. • New nuclear units eliminate natural gas builds and some renewables which were the primary expansion units in the reference case. Energy Efficiency levels are similar to the reference plan (case 1A). • System-wide CO2 emissions are lower as the generation from the nuclear units replaces gas generation and displaces existing coal generation.

Renewable Sensitivities • Extension of Solar Tax Credits • Higher HVDC Wind Net Dependable Capacity & lower cost • Slower Solar Cost De-escalation • Slower Wind Cost De-escalation Resource Sensitivities • Pumped-hydro storage • Compressed air energy storage • Integrated gas combined cycle with carbon capture and sequestration • Supercritical pulverized coal 1x8 with carbon capture and sequestration • New direct combustion Biomass

Energy Efficiency & Demand Response Sensitivity Results: These cases examined several key Energy Efficiency inputs, including testing the impact of removing the planning factor adjustment (discussed in Appendix D) and accelerating the near-term ramp rates for Energy Efficiency. Cases were also run with no Energy Efficiency or Demand Response in the portfolio. Conclusions are as follows:

Key Driver Sensitivities • Higher load • No CO2 penalty • Lower gas price • Higher gas price

• Removing the planning factor adjustment results in similar Energy Efficiency volumes as the reference case (case 1A) through 2023, increasing thereafter to midway between the reference case and the Maximize Energy Efficiency strategy (case 1D) by 2033.

Figure 8-10: List of Sensitivity Cases

108



• Increasing the ramp rate in the early years of the study results in small increases in Energy Efficiency by 2033 with slightly more selections near to midterm and little impact to overall system cost. • Higher volumes of Energy Efficiency equate to higher system average costs because electricity sales are lower. There is a tradeoff between average system cost and total system costs even in the reference case. • Energy Efficiency continues to be perform as a resource in model results: ‐‐ Energy Efficiency programs eliminate some of the need for natural gas units as well as some renewable purchases. Energy Efficiency volumes reduce generation from gas, coal, and renewable resources. ‐‐ Demand Response programs eliminate some of the need for natural gas peaking units and market purchases.

Resource Sensitivity Results: This group of cases examined the impact of forcing in certain resource types not selected in the original case runs. These include forcing in pumped storage, compressed air energy storage, pulverized coal with carbon capture and sequestration (CCS), integrated gasification combined cycle (IGCC) with CCS and biomass plants. These resources were added in 2028 and the portfolio was re-optimized. Conclusions are as follows: • Coal options generally displace demand response, natural gas and renewable generating assets. Each coal option increases total system costs. • Biomass options generally offset small amounts of demand response. • Pumped storage generation offsets future gas generation and some renewables. • Compressed Air Energy Storage generally offsets gas peaking and demand response assets.

Renewables Sensitivity Results: These sensitivity cases examined the impact of lower cost assumptions for wind and solar resources and more favorable guaranteed capacity from High Voltage Direct Current (HVDC) wind. Cases were also analyzed that assumed solar and wind costs do not decline as quickly as assumed in the original scenarios. Conclusions are as follows:

Key Driver Sensitivity Results: These cases analyzed the impact of changing one key driver while leaving other case inputs unchanged. A more aggressive load growth case was tested, as were higher and lower gas prices. In addition, a case with no CO2 cost was analyzed. Conclusions are as follows: • If loads are materially higher than expected, resource needs are primarily met with new natural gas builds and market purchases; renewables and EE remain similar to reference case. • In a low gas price case, more natural gas units are built and additional coal is retired. There are also fewer renewable purchases and less energy efficiency. • In a high gas price case there are fewer natural gas units built, more renewable purchases and more coal generation. • If the CO2 price penalty is removed there is additional coal generation and fewer renewable purchases.

• Lowering costs and providing a higher guaranteed net dependable capacity for HVDC wind results in selection as early as 2020. • Assuming lower prices driven by tax policy and availability of favorable solar sites, utility-scale solar tracking is selected as early as 2020. • Increasing solar escalation rates pushes out utilityscale solar selection to 2029 and halves the volume compared to reference case. • Increasing wind escalation rates pushes out wind selection to beyond 2033. • As seen in other sensitivity cases, renewable selection is highly sensitive to gas prices.

109



Overall Conclusions: The sensitivity case runs were instrumental in informing TVA and the stakeholder working groups about key inputs affecting the results. In general, the sensitivity cases confirmed that our original case study results formed a reasonable boundary of future resource additions. Other high-level conclusions are summarized below: • New nuclear or coal assets would offset gas builds and renewable purchases, increase total cost and lower fuel risk. Cost-sharing would render SMRs more attractive, and nuclear additions may prove more valuable in future IRPs given their long lives and the possible expiration of some of our existing nuclear licenses that may occur just beyond the study window. • The original EE case results (strategy D) still form an effective boundary for EE results, and energy efficiency programs eliminate the need for most natural gas builds and some renewable purchases. Removing the planning factor adjustment does not affect the near-term selection of energy efficiency and results in selections midway between case 1A and 1D at the end of the study period. Increasing near-term ramp rates does not materially change the overall trajectory or costs. Finally, higher volumes of EE result in higher system average costs, even in the Reference Plan (case 1A), driven by lower electric sales. In some cases, the impact to average cost is similar to the impact of adding new nuclear builds to the portfolio. • Renewable selection is highly dependent on gas price assumptions, load, and unit cost and characteristics. • Natural gas pricing remains a key sensitivity for all resource selections.

110

Chapter 9

Recommendations Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

111

Re-evaluate

Recommend


Chapter 9: Recommendations

9 Introduction

9.1 Study Objectives

The Tennessee Valley Authority’s 2015 Integrated Resource Plan (IRP) will guide TVA in making decisions about the energy resources used to meet future demand for electricity. Having a long-range resource plan enables us to provide affordable, reliable electricity to the people we serve. It is a crucial element for success in a constantly changing business and regulatory environment and will better equip us to meet many of the challenges facing the electric utility industry in the coming years.

The following objectives guided the development of the IRP: 1. Deliver a plan aligned to least-cost planning principles. 2. Manage risk by utilizing a diverse portfolio of supply and demand-side resources. 3. Deliver cleaner energy and continue to reduce environmental impacts. 4. Evaluate increased use of renewables, energy efficiency, and demand response resources. 5. Ensure the portfolio delivers energy in a reliable manner. 6. Develop the ability to dynamically model energy efficiency in the study. 7. Provide flexibility to adapt to changing market conditions and identify significant sign posts. 8. Improve credibility and trust through a collaborative and transparent process. 9. Integrate stakeholder perspectives throughout the study.

We used an integrated, least-cost system planning process that takes into account the demand for electricity, resource diversity, reliability, costs, risks, environmental impacts, and the unique attributes of different energy resources. Various ranges of inflation, commodity prices, and environmental regulations were evaluated to provide needed information. Constraints, including corporate, strategic, and environmental objectives, were considered as different combinations of strategies and predictions of future conditions were analyzed and evaluated. We conducted the IRP process in a transparent, inclusive manner that provided numerous opportunities for public education and participation. The analysis performed in this IRP study relied on industry-standard models and incorporated best practices while using an innovative methodology to more fully evaluate the role of energy efficiency and renewable resources in the power supply mix. Resource cost and performance input data were independently validated.

The analysis performed in the study is intended to identify a resource mix that positions TVA for success regardless of how the future unfolds. The resulting power supply mix will meet these goals: low cost, reliable, risk-informed, diverse, environmentally responsible and flexible.

9.2 Findings The IRP study demonstrates that TVA power will continue to be reliable, affordable and sustainable into the future. Our resource additions will build on TVA’s existing diverse asset portfolio reflected in Figure 9-1.

112



Figure 9-1: 2014 TVA Portfolio

Study results show that there is no immediate need for new base load plants after Watts Bar Nuclear Unit 2 comes online and uprates are completed at Browns Ferry Nuclear Plant. Instead, we can rely on additional natural gas generation, greater levels of cost-effective energy efficiency and increased contributions from competitively priced solar and wind power. We also expect to have less coal-based generation in our energy mix than we do today. In all cases, TVA will continue to provide for economic development in the Tennessee Valley.

footprint and position the Tennessee Valley to have significant reductions in CO2 emissions. Strategies that emphasize energy efficiency or renewables have the best environmental results. • Valley Economics: All strategies have a similar impact on overall economic health and contribute to a strong, vibrant economy across the region. • Flexibility: System flexibility is generally equivalent in most cases but is reduced when renewables are strongly emphasized. Reviewing these results led to questions from stakeholders about how changes in assumptions or resource choices might impact the findings. A series of sensitivity cases were evaluated with five key assumption categories: nuclear additions, modified assumptions for energy efficiency, alternative renewable resource costs, impacts associated with forcing resources not otherwise selected into the mix, and changes in fundamental drivers such as load growth and fuel pricing. The results of these analyses supported the ranges established in the initial findings. The sensitivity cases, coupled with the original 25 case results, provide a robust set of potential resource additions evaluated in the IRP from which the final recommendations were derived. Figure 9-2 provides the range of capacity additions (by 2033, rounded) from the IRP case and sensitivity analysis:

We identified five key measures to evaluate the performance of the plans created as part of the study. A review of the case results produced the following outcomes: • Cost: Total costs are similar for many of the cases over the long term, and strategies that allow for a diverse mix of resource additions have a lower cost than those that emphasize particular technologies. Higher amounts of energy efficiency may create an upward pressure on rates in future years due to reduced sales. • Financial Risk: Risks are minimized by maintaining a diverse portfolio and not over-emphasizing any specific resource type. • Environmental Stewardship: All strategies show significant improvement in TVA’s environmental

113



Figure 9-2: Evaluated MW Additions/Retirements by 2033 from IRP Base and Sensitivity Case Analysis Key findings from the sensitivity cases are summarized below:

• Renewable selection is highly dependent on gas price, load, and cost and performance assumptions. • Natural gas pricing and load levels remain key sensitivities for all resource decisions.

• New nuclear or coal assets would offset gas builds and renewable purchases. Nuclear additions increase total cost but lower fuel risk. Small Modular Reactors are presently cost-prohibitive, but cost-sharing would render them more financially attractive. Subsequent IRPs will need to address the expected expiration of licenses for TVA’s operating nuclear units which may occur beyond the present study window. • Energy efficiency was successfully modeled as a selectable, supply-side equivalent resource. In general, energy efficiency programs eliminate the need for natural gas units as well as some renewable purchases. As with any resource, cost and performance assumptions are critical to the final result, and lower costs or uncertainty around this resource would increase its selection in the portfolio. Our study results also highlight that higher volumes of energy efficiency tend to increase system average costs. TVA and our local power company partners will need to balance energy efficiency volumes and programs to ensure that those who cannot participate in these programs are not disproportionately impacted.

9.3 Developing the Recommendation The recommendation takes into account customer priorities around power cost and reliability across different futures. Implementing the least-cost resource plan with these priorities in mind will help ensure TVA continues to fulfill its mission to serve the people of the Tennessee Valley. In developing a recommendation from the study, TVA has elected to establish guideline ranges for key resource types (owned or contracted) that make up the target power supply mix. This general planning direction is expressed over the 20-year study period while also including more specific direction over the first 10-year period. In order to distill the considerable number of cases evaluated through the original scenario and strategy analysis and the sensitivity cases, the recommendation uses ranges that are centered on results obtained under the Current Outlook scenario. The other scenario results provide a sense of how the recommended mix might change as the future changes.

114



The need to shift the resource mix will be based on these key variables:

the end year of the study (2033), shown in Megawatts (MW). The results are drawn from strategies A, B and C which do not place specific targets on particular resource types. Strategies D and E were intentionally designed to focus on meeting future resource needs with certain resource types only (Energy Efficiency and Renewables). The results of these strategies are included in the analysis but are not part of the recommended range because of their limited focus on particular resource types. These strategies provided valuable insight into the planning process and provide outer bounds to which TVA could navigate if certain developments or conditions unfold.

• Changes in the load forecast. • The price of natural gas and other commodities. • The pricing and performance of energy efficiency and renewable resources. • Impacts from regulatory policy or breakthrough technologies. The first three variables represent the fundamental drivers for most of the variation in the resource plans produced across the strategy/scenario combinations. Our planning direction, while initially focused around the current view of the future, is flexible enough to indicate how that power supply mix would shift if one or more of these key variables exhibits a material change from the forecasts used in the IRP. We also recognize that impacts from breakthrough technologies (like a significant advance in energy storage) would be a game-changer, and TVA will continue to monitor emerging technology as it develops.

The solid bars represent the range of results from strategies A-C in the Current Outlook scenario, which represents our best estimation of the future. However, recognizing that a variety of future scenarios are possible, we provide a broader range (shown in horizontal black lines) to represent how the resource portfolio may respond in different future scenarios. The range for Energy Efficiency and Demand Response also incorporates TVA’s current trajectory for these resources to account for some of the implementation and policy uncertainties discussed in Appendix D and Chapter 10.

This approach provides a more robust recommendation than was developed in the 2011 IRP. While that approach provided a solid framework for the resource decisions TVA has made since the TVA Board accepted the IRP planning direction in the spring of 2011, the changing utility marketplace requires a more flexible guide that provides decision-makers with a clear understanding of how the resource mix would evolve in response to future uncertainties. The recommendation meets the dual objective of ensuring flexibility to respond to the future while providing guidance on how our resource portfolio should change as the future unfolds.

The recommended ranges represent incremental additions (or retirements) from the existing resource fleet and include contracted (market) positions that can be sourced from resources that meet cost and performance requirements, providing flexibility for the portfolio. The results are bounded by the full range of the IRP cases and sensitivity runs which affirm the merits of a diverse portfolio. TVA will closely monitor key input variables including changes in the load forecast, the price of natural gas and other commodities, the pricing and performance of energy efficiency and renewable resources, and impacts from regulatory policy or breakthrough technologies to help determine whether adjustments should be made to the recommended ranges.

9.4 Target Power Supply Mix The recommendations for the power supply mix are presented in the form of ranges around boundaries established by the IRP results. Figure 9-3 shows the range of resource additions we are proposing by the end of the first 10 years of the study (2023) and by

115



Figure 9-3: Range of MW Additions by 2023 & 203315

Recommendations by resource type: Coal: Continue with announced plans to retire units at Allen, Colbert, Johnsonville, Paradise and Widows Creek. Evaluate the potential retirement of Shawnee Fossil Plant in the mid-2020s if additional environmental controls are required. Consider retirements of fully controlled units if cost effective.

Demand Response: Add between 450 and 575 MWs of demand reduction by 2023 and similar amounts by 2033, dependent on availability and cost of this customer-owned resource. Energy Efficiency: Achieve savings between 900 and 1,300 MW by 2023 and between 2,000 and 2,800 MWs by 2033. Work with our local power company partners to refine delivery mechanisms, program designs and program efficiencies with the goal of lowering total cost and increasing deliveries of efficiency programs.

Nuclear: Complete Watts Bar Nuclear Unit 2 and pursue additional power uprates at all three Browns Ferry units by 2023. Continue work on Small Modular Reactors as part of technology innovation efforts and look for opportunities for cost sharing to render these more cost-effective for our ratepayers.

Solar: add between 150 and 800 MW of large-scale solar by 2023 and between 3,150 and 3,800 MW of large-scale solar by 2033. The trajectory and timing of solar additions will be highly dependent on pricing, performance and integration costs.

Hydro: Pursue an additional 50 MW of hydro capacity at TVA facilities and consider additional hydro opportunities where feasible.

15 MWs are incremental additions from 2014 forward to align to the IRP analysis base year. Board-approved coal retirements and natural gas additions as of August 2015 are excluded from the totals.

116



Wind: Add between 500 and 1,750 MW by 2033, dependent on pricing, performance and integration costs. Given the variability of wind selections in the scenarios, evaluate accelerating wind deliveries into the first 10 years of the plan if operational characteristics and pricing result in lower-cost options. Natural Gas (Combustion Turbine and Combined Cycle): Add between 700 and 2,300 MW by 2023 and between 3,900 and 5,500 MW by 2033. The key determinants of future natural gas needs are trajectories on natural gas pricing and energy efficiency and renewables pricing and availability. TVA’s recommended planning direction affirms its commitment to a diverse resource portfolio guided by the least-cost system planning mandate. The ranges above provide a general guideline for resource selection, but the full case analysis studied in the IRP and the SEIS includes ranges much broader than shown above driven by key drivers such as significant changes in economic conditions or regulations. We believe meeting our future needs in accordance with the resource technologies and ranges in this recommendation will position TVA to continue to deliver reliable, low-cost, and cleaner power to the people of the Tennessee Valley.

117

Chapter 10

Implementation Challenges and Next Steps Scoping

Inputs & Framework

Analyze & Evaluate

Present Findings

119

Re-evaluate

Recommend


Chapter 10: Implementation Challenges and Next Steps

10 Implementation Challenges and Next Steps

knowledge about technology advances and changing market conditions; (2) TVA should work closely with local power companies as energy efficiency efforts and distribution-level resources are implemented; and (3) TVA and appropriate partners should investigate additional approaches in energy efficiencies and distributed resources, considering those who cannot afford the necessary investments and recognizing fairness and equity for all rate payers.

This chapter outlines some of the challenges TVA faces in implementing the recommendations of the IRP study and discusses key policy considerations and improvements to modeling and the study process.

10.1 Overview of Next Steps In the Draft Report, we provided a high-level schedule for next steps that included the release of the draft IRP/ SEIS reports and the completion of a public comment period. Now that the comment period is concluded, we have moved to the final two steps in our IRP process as shown in Figure 10.1: Spring/Summer 2015

Incorporate Input Review comments Complete additional analyses if needed • Revise the study report • •

Implementing the recommendations from the IRP will require close cooperation between TVA, local stakeholders, our Local Power Company (LPC) partners and Valley electric customers particularly around deployment of additional energy efficiency resources. The success of energy efficiency depends on end-use customer participation. TVA is primarily a wholesale power provider and the LPCs have the relationship with most end-use customers. TVA will need to partner with LPCs and others in the region to design additional delivery mechanisms to achieve the levels of penetration envisioned in the IRP. We have a history of successful collaboration around the design and delivery of EE programs and plan to build on that experience. There are a number of initiatives already underway both internal to TVA and in cooperation with our LPC partners seeking more effective and innovative program designs, improved performance tracking and budgeting, and enhanced delivery mechanisms.

Summer 2015

Identify Target Power Supply Mix Develop study recommendations • Prepare final report and post • Request TVA Board action •

Figure 10.1: Remaining IRP Process Steps

Similarly, TVA’s status as a power wholesaler complicates deployment of cost-effective renewable resources (primarily solar). The IRP envisioned using utility-scale solar resources which can be located to provide the most value to the transmission or distribution systems. While TVA owns and operates the high-voltage transmission grid, the distribution system is actually a grid of grids belonging to the 155 LPCs, each with its own unique characteristics and operational challenges. Renewable resources installed on the distribution grid necessitate the involvement of entities in addition to TVA, especially the LPCs. This is especially true for small-scale distributed (rooftop) solar resources. Although TVA did not include small-scale rooftop solar as a resource option in the IRP, we did include small-

After TVA issues the final IRP and SEIS, there is a 30-day waiting period before the TVA Board of Directors can be asked to make a decision about the IRP. After the Board makes a decision, the NEPA process is completed by issuing a Record of Decision that documents the Board’s action and its basis.

10.2 Implementation Challenges The Regional Energy Resource Council (RERC), after reviewing the recommendations in the IRP, offered the following advice to the TVA Board focused on implementing those recommendations: (1) TVA should consider all of the analyses in the IRP and continue to refine input assumptions based on relevant data,

120



scale commercial solar as an option, and we analyzed significant levels of distributed generation penetration in the scenarios to help us begin to understand how the increasing use of distributed generation will affect the TVA power system.

consequences for low/fixed income customers as well as renters or other customers that do not participate in the programs, which doesn’t fit with our mission. IRP analyses we have completed will help inform the followon planning and evaluation of the EE portfolio. TVA recognizes that EE should be a key part of our power supply mix consistent with the findings in the IRP. We also know that program design will be a key challenge to ensure that the broadest possible EE portfolio can be offered through the LPCs to minimize possible bill impacts on non-participants.

TVA is leading an initiative with the goal of determining the value of distributed resources on the system. Initial efforts are focused on small-scale distributed (rooftop) solar, but the method is general enough to allow for value determination for other distributed options. Work is ongoing, led by a team that includes technical support from the Electric Power Research Institute (EPRI), to develop a methodology to identify site preferences on the distribution systems of the LPCs. This work, along with locational analysis already completed by TVA, will help in placement of utilityscale and distributed solar in support of the IRP recommendations.

We also realize that the level of electric rates and job growth are critical concerns for Valley governments, businesses and residents. The IRP uses two specific metrics for the macro-economic impacts of resource strategies. These metrics and underlying analyses provide important information about future revenue requirements that affect future rate levels and will help inform the future direction of TVA’s economic development program. However, none of the strategies had a significantly different impact from the others on the Valley economy. Section 7.5.7 of the SEIS provides more information about socioeconomic effects.

Finally, it is important to note that the recommendations in the IRP also include more traditional resources, primarily gas-fired, that come with their own implementation challenges in the areas of siting and permitting both for the units themselves and for the transmission lines and gas pipelines associated with them. TVA has several teams working on various aspects of the siting and permitting work necessary to ensure that when these resources are needed as part of the generation portfolio, we will be better positioned to bring add them to the resource mix.

There are several other policy issues that come into play when implementing recommendations from the IRP. For example, we know that the EPA’s Clean Power Plan will be finalized at virtually the same time this report is released. We will look at that final rule more specifically to understand how the IRP can inform TVA’s compliance plans, but feel the study recommendations point us in a direction to meet whatever requirements are included in that rule. Because of our unique business model, TVA, stakeholders, electric customers and its Local Power Company partners will have to collaborate in new and innovative ways to ensure that this evolving resource portfolio remains reliable and provides maximum value to all customers.

10.3 Policy Considerations The IRP is a resource planning study focused on identifying a target power supply mix for TVA. In the process of developing the cases and reviewing the results with stakeholders, a number of policy-related issues were raised that are outside the scope of the IRP itself but will need to be considered as we move toward implementation of recommendations from the study.

10.4 Process and Modeling Improvements

For example, we recognize that a commitment to significant levels of energy efficiency as part of the resource portfolio will likely put upward pressure on rates (absent a redesign), and that could have negative

As this IRP cycle winds down, we anticipate undertaking the next study within five years, sooner in this period if one or more of our indicators trigger the

121



analysis. As part of the after-action review of this study we have identified a number of improvements to either the study process or the models we use to conduct the planning study. These include continuing to develop the modeling approach to treating EE as a selectable resource, enhancing our process for developing the scenario/strategy framework and reviewing and refining the IRP process to ensure strong stakeholder feedback remains a key component of the study.

10.5 Conclusion TVA finds considerable value in undertaking an IRP and especially appreciates the input, review and insights of individuals on the IRP Working Group and the Regional Energy Resource Council. They spent considerable time helping us develop a robust plan that meets all the criteria outlined in our objectives. TVA values their involvement and expertise on behalf of all our stakeholders in making this a better IRP. As with any long-term resource plan that attempts to prepare for the future, TVA’s IRP reflects what we know today and can reasonably expect for the coming years. TVA, and our employees across the Valley, stand ready each and every day to continue our three-part mission around energy, the environment and economic development. We will do our best to continue to serve the people of the Tennessee Valley by providing low cost, reliable power in an environmentally responsible manner while promoting economic development across the Valley.

122

Appendix A

Generating Resource Cost and Performance Estimates

123


Appendix A

A wide array of new resource options were available in the capacity planning expansion models for selection to meet load growth or fill resource needs. Each new resource option has a set of unique characteristics such as capacity, construction time, book life, heat rate, outage rate, capital cost, variable cost, and fixed costs. Chapter 5 includes a discuusion of the resource options considered in the IRP. An independent thirdparty, Navigant Consulting, Inc. (Navigant), reviewed and compared the TVA planning parameters used in the IRP to proprietary and other industry sources to ensure the modeled unit characteristics and assumptions were representative of the respective generating technologies. This appendix contains aletter report summarizing the benchmarking efforts of Navigant as well as TVA’s internal benchmarking efforts.

independently verified for either accuracy or validity, and no assurances are offered with respect thereto. This Report does not represent any endorsement of any particular resource type, nor a guarantee that any resource type is viable or can be ultimately delivered. This Report covers the TVA 2015 IRP only. We make no representations, warranties or opinions concerning the enforceability or legality of the laws, regulations, rules, agreements or other similar documents reviewed as part of this work. Navigant and its employees are independent contractors providing professional services to TVA and are not officers, employees, or agents of TVA. Background and Scope As part of the 2015 IRP effort, TVA is identifying and evaluating potential new power generating and storage resources necessary to serve future load. Estimated values for new resource cost and performance are necessary in order to perform generation capacity expansion and dispatch modeling. TVA requires estimated values that are internally consistent and representative of actual values to be observed in practice. Parameters include performance and cost for traditional, renewable, and alternative generation technologies, and also for power storage technologies. Estimated values are obtained from several sources including the TVA business units, the Tennessee Valley Renewable Information Exchange, and the IRP project staff itself.

A.1 summary Letter: Navigant Benchmarking Report Summary Letter Report on Generating Resource Cost and Performance Estimates Developed for the 2015 TVA Integrated Resource Plan September 12, 2014 Navigant Consulting, Inc., (“Navigant”) has reviewed and recommended cost and performance parameters for potential new power generation and storage resource alternatives to be considered in the Tennessee Valley Authority (TVA) 2015 Integrated Resource Plan (IRP) (“Resource Estimates”). The work was performed for TVA under contract work authorization #669468 and purchase order #709838 (revised). The primary deliverable was a Microsoft Excel workbook summarizing the Resource Estimates and related assumptions and notes. The preliminary draft workbook was delivered on April 25, 2014, and the final workbook was delivered on June 17, 2014.

Navigant’s task was to review the estimated values provided by TVA for each resource type, and, as necessary, develop alternative values, forming a set of Resource Estimates that are indicative of what can be expected for each resource technology within the Tennessee Valley geographic area. The deliverable was a spreadsheet workbook of tables – one for each resource technology – that:

This report (“Report”) summarizes the work scope, the resources and parameters reviewed, and our primary findings at a high level. In performance of this review and Report, we have in part relied on information provided to us by TVA and third parties. While we believe this information to be reliable, it has not been

• lists the parameters and associated values provided by TVA, • lists alternative values as available and relevant, and • recommends specific Resource Estimates for use in IRP modeling.

124


Appendix A

outage rate, storage efficiency, storage input demand, plant overnight capital cost, transmission upgrade cost, total overnight capital cost, variable operating & maintenance (O&M) cost, fixed operating & maintenance cost (both in $ and $/kW-year), firm gas charge, and book life.

Technologies and Parameters Reviewed Power generation and energy storage resources considered in the review included the following, which represent alternatives for new capacity to serve future load: • Natural gas-fired generation ‐‐ Single cycle combustion turbines ‐‐ Combined cycle combustion turbines (with and without supplemental duct firing) • Coal-fired generation ‐‐ Pulverized coal (with and without carbon capture and sequestration) ‐‐ Integrated gasification combined cycle (coal) (with and without carbon capture and sequestration) • Nuclear generation ‐‐ BW205 design ‐‐ AP1000 design ‐‐ Small modular reactors • Energy storage ‐‐ Pumped hydro-electric storage ‐‐ Compressed air energy storage (CAES) • Solar photovoltaic (PV) generation ‐‐ Utility scale (both fixed-panel and tracking) ‐‐ Commercial scale (both small and large) • Wind energy generation ‐‐ Onshore within the Tennessee Valley ‐‐ Located in Midcontinent Integrated System Operator (MISO) or Southwest Power Pool (SPP) ‐‐ Obtained via High Voltage Direct Current (HVDC) transmission • Biomass energy generation ‐‐ Co-firing ‐‐ Integrated gasification combined cycle (IGCC) (biomass) ‐‐ Direct combustion at new facility ‐‐ Repowering of existing facility

When relevant and reliable industry values for specific parameter values were available, they were utilized for comparison and as a basis for any Resource Estimate. Notes concerning the source and reconciliation of any material differences were provided in the workbook. High-Level Findings and Recommendations Navigant provided recommended parameter values and performed direct comparisons with TVA estimates for 264 values. For about two-thirds of these, the TVA values were determined to be consistent with the recommended values (meaning within 10%, measured relative to the original TVA estimate). The remaining one-third of the values showed numerical differences of greater than 10%, characterized here as “material”. Of the materially different values, over half – representing 62 of the 264 values reviewed – showed differences greater than 20%. Some parameters are correlated with others, and one key difference in interpretation or estimation sometimes led to a pattern of differences across parameters. Additionally, variations in underlying classification categories (cost allocation, for example) can mean that there is some compensation or offsetting in net effects when modeling. Overall, the substantial majority of TVA values were determined to be consistent with recommended values, and otherwise reasonable. Regarding natural gas-fired generating resources, for the 48 parameter values compared, 29 (59%) of the TVA values were consistent with values recommended by Navigant. Roughly one-fifth of all parameters showed differences of 20% or more. The only systematic material difference between TVA values and recommended values was in annual outage rates, where the Navigant recommendations were higher across the board. For a given resource, parameter value differences vary in terms of impact, and a number of

Cost and performance parameters vary somewhat according to generating and storage technology, but each technology generally has 8-12 applicable characteristics or parameters for which values were reviewed. These include summer net dependable capacity, summer full-load heat rate, build time, annual

125


Appendix A

potentially offsetting differences are evident.

located for biomass IGCC, and there are no such plants in service.) Where comparisons were possible, recommendations were materially higher for heat rate, build time, outage rate, and plant overnight capital. Some differences were due to varying assumptions about plant sizing, however, and some potentially offsetting differences were noted for variable and fixed O&M.

The vast majority (79%) of the 66 coal resource parameters compared were in agreement. For the parameters with material differences, there was no systematic pattern, although some differences were noted for plant overnight capital costs, build time, and variable O&M. For nuclear generation, about half of the parameter values (15 out of 31) were found to be consistent. Most of the remaining values were 20% or more different (12 values). Generally speaking, recommended outage rates, plant and total overnight capital costs, and variable O&M values were materially higher than TVA values.

On balance for all the generating and storage resources examined, the substantial majority of the proposed TVA parameter values for which comparisons were performed were consistent with recommended values – about two-thirds of all compared values. For those parameters with material differences in values of 10% or more, a number of those were to some degree offsetting within a given resource/technology.

Regarding energy storage, two-thirds of the compared parameter values were materially consistent. Each value with a material difference was at least 20% different. The parameters with such differences included variable O&M, fixed O&M (both dollars per year and $/kw-year), and book life for pumped hydro; and annual outage rate, storage efficiency, and plant and total overnight costs for CAES. Some potentially offsetting differences were observed.

The TVA values reviewed were provided in spring 2014, and the summary above relates to recommendations and comparisons based on the values provided at that time. Since then, TVA has modified numerous values to be used in its IRP modeling, in part reflecting the outcome of this review. TVA staff was extremely helpful and responsive both in providing supporting information needed in the review/comparison process, and in providing useful feedback and clarification on the draft workbook deliverable and the constituent parameter values. It is clear that TVA is striving to fairly represent all of the potential new generating resources in its IRP modeling, thus laying the basis for meaningful IRP modeling of resource expansion alternatives.

Almost all of the solar PV parameter values compared were consistent. Only a single material difference was identified, where the recommended value for fixed O&M (small commercial rooftop solar) was materially higher. For wind energy, 16 of the 29 parameter values compared (or 55%) were consistent, with about half of the remaining values showing differences greater than 20%. Recommended outage rates were materially higher than TVA values for all three technology alternatives. Other differences varied by technology, and some potentially offsetting effects are seen.

A.2 TVA Benchmarking summary: Optimizing Asset Decisions When evaluating how to best meet future customer needs for electricity, TVA optimizes decisions using least-cost planning models. These models require inputs on variables such as capacity amounts, upfront capital costs, and fuel usage parameters, and many others. The models integrate all the variables for new resources under the various scenarios (i.e., various fuel prices, demand projections, regulatory environments, etc) to select expansion units that best fit the portfolio needs and requirements in a total least-cost manner.

Biomass options show consistent parameter values in about one-quarter of the comparisons, with material differences in about three-quarters of the 29 values compared. All of materially different values are at least 20% different. This applies to co-firing, new direct combustion, and biomass repowering of existing coal. (No reliable source of industry information was

126


Appendix A

One of the key assumptions that contributes to resource selection is the cost to construct a particular unit. Construction and capital costs are determined from industry experience, vendor information, benchmarking, et cetera. These costs are presented as Overnight Capital Costs in the table. This is the cost to build the asset and is computed as total dollars divided by the capacity of the unit in kilowatts ($/kW).

Supply Option1

Unit Characteristics Capacity (MW)

Natural Gas

Coal

Depending on how an asset’s dispatch cost compares to other assets in the fleet, the amount of energy sourced from an asset may vary greatly over time. For example, when natural gas prices are low, those assets powered with natural gas serve customers with more energy than when natural gas prices are high. A concept that is sometimes used to compare asset costs is LCOE or Levelized Cost of Energy. This measure divides the total cost of an asset (i.e., construction and capital, ongoing maintenance and operating, and dispatch costs which are primarily fuel) by expected output or generation.

Nuclear3 Storage

Solar4

Wind4

Biomass Hydro

Because dispatch costs and expected output vary widely across all of the IRP scenarios, LCOE is not a useful metric to benchmark resource costs. A better comparison, and the standard for resource planning, is to compare $/kW installed capital costs. These are the actual inputs into the capacity expansion model and the costs benchmarked by TVA’s independent third-party contractor.

Energy Efficiency

CT 3x CT 4x CC 2 by 1 CC 3 by 1 IGCC PC 1x8 PC 2x8 IGCC_CCS PC 1x8_CCS PC 2x8_CCS PWR APWR SMR Pump Storage CAES Utility_Tracking Utility_Fixed Commercial_Small Commercial_Large MISO SPP In Valley HVDC Direct Combustion Repowering Existing Spill Addition Space Addition Run of River Res Tier 1 Res Tier 2 Res Tier 3 Comm Tier 1 Comm Tier 2 Comm Tier 3 Ind Tier 1 Ind Tier 2 Ind Tier 3

590 786 670 1,005 500 800 1,600 469 600 1,200 1,260 1,117 334 850 330 25 25 25 25 200 200 120 250 115 75 40 30 25 10 10 10 10 10 10 10 10 10

Overnight Capital Cost2 (2013$/kW) $738 $712 $1,097 $1,030 $3,845 $2,908 $2,722 $7,286 $6,518 $6,271 $5,460 $5,856 $8,252 $2,365 $1,072 $1080-$2353 $1080-$2059 $3,529 $2,941 $1,750 $1,750 $1,875 $1421-$2242 $4,357 $1,092 $2,200 $1,800 $2,550 $2,076 $2,911 $3,817 $1,168 $1,931 $3,341 $1,154 $1,908 $3,302

Footnotes: 1. Supply options represent generic site build costs, except the PWR resource which represents the Bellefonte site option. 2. Overnight capital costs do not include Allowance for Funds Used During Construction (AFUDC). All options include a generic transmission upgrade costs. 3. The PWR and APWR costs are for the first unit. The SMR cost is for a twin pack. 4. The capital costs for solar and wind assume that tax credits expire/decrease per current federal law. Sensitivity cases on utility tracking and HVDC wind test impact of extensions (range reflects capital cost range for sensitivity analysis). Solar capital costs are assumed to decline over time per recent trajectories and wind capital costs increase at less than the rate of inflation.

Benchmarking Capital Costs TVA engaged an independent third-party, Navigant Consulting (NCI), to review cost and performance assumptions proposed for use in the 2015 IRP. NCI evaluated our assumptions for various unit types along with cost assumptions for renewable resources developed in a collaborative effort with stakeholders. This independent assessment found that the majority of assumptions proposed for the study were consistent with typical values used in the industry. Many of the remaining assumptions were modified based on NCI recommendations prior to running the IRP cases. The data in the table presented in the preceding section reflect adjustments recommended by NCI.

We have also prepared a comparison of our capital cost assumptions from the IRP study to a recent Lazard report released in September 2014 to further demonstrate the reasonableness of those assumptions. The capital cost data from the summary table has been adjusted to match assumptions used in this Lazard report (expressed in comparable year and including financing costs). This comparison chart shows how TVA’s assumptions on capital cost compare to those

127


Appendix A

recently published by Lazard. The cost comparisons are generally consistent given Lazard’s study is based on nation averages and TVA’s costs are specific to the TVA

system. In addition, footnotes are provided to explain variations for each asset type.

* Source: Lazard’s Levelized Cost of Energy Analysis – Version 8.0. September 2014 ** Source: TVA’s 2015 Integrated Resource Plan. Assumptions are in 2015$/kW capital costs and include financing costs. 1. The high end range for TVA represents a small commercial solar unit and the low end represents a large commercial solar installation. 2. Lazard’s high end represents a solar tracking unit and the low end represents a fixed-tilt solar unit. TVA’s low end represents the solar sensitivity unit and the high end represents the solar tracking unit. Solar capital cost assumptions decline at 3.5% a year through 2020 and then remain flat through 2029. 3. TVA’s high end represents the HVDC option; the low end represents the wind sensitivity analysis. These costs do not include transmission wheeling charges (similar to Lazard’s). 4. The TVA biomass unit assumptions are modeled after a recent utility-scale plant. The basis for the Lazard assumption is not specified. 5. The high end costs include carbon capture and storage. 6. The low end represents PWR (BLN), the midpoint is an AP1000, and the high end represents an SMR.

128

Appendix B

Assumptions for Renewables

129


Appendix B

Modeling Approach for Wind & Solar Options

factors. For modeling purposes, TVA assumed the MISO and SPP wind had a 40 percent capacity factor, the HVDC option originating from Oklahoma had a 55 percent capacity factor, and the In-valley option had a 30 percent capacity factor.

Wind and solar resources have unique operating characteristics that are different from thermal and other more traditional resources. To properly account for the contribution from these intermittent resources, the energy contribution is represented using hourly energy profiles that are imported into the model, and the seasonal capacity of these resources is represented by a computed Net Dependable Capacity (NDC) value. The annual capacity factor of the hourly energy profiles are also computed to ensure the total amount of energy is comparable to industry benchmark sources. This appendix discusses the methodology TVA has used to determine both the energy profiles and NDC values for wind and solar options that are considered in the IRP.

The HVDC project has a 55 percent annual capacity factor due to the availability of wind in Oklahoma and the newer technology of the wind turbines, which were assumed to be GE 1.7-100 wind turbines at a height of 80 meters. This capacity factor is much higher than TVA’s existing wind contracts in other locations. The chart below shows the range of capacity factors: Wind Capacity Factors

HVDC (future) MISO (future) SPP (future)

Wind Modeling Generation from wind is weather and location dependent, and not dispatchable like more conventional resources. Therefore, utilities need to develop a reasonable representation of the output from wind for use in long-range planning models. This “wind shape” is based on actual data collected from specific sites, or modeled data using wind turbine design assumptions.

In Valley (future)

20% 0%

10%

20%

46% 42% 40% 40% 40% 38% 37% 37% 36% 35% 32% 30% 30%

40%

50%

55%

TVA PJM SPP MISO

60%

Determining the Wind Net Dependable Capacity (NDC)

TVA uses data from 3TIER to develop the planning assumptions around wind shape and capacity factor for use in the IRP. A “typical week” hourly shape for each month was developed by 3TIER for each wind option. Once a shape has been selected, the amount of energy produced can be determined and a capacity factor computed (actual generation expressed as a percentage of maximum possible generation).

Planners must determine how much wind generation is likely at the system peak hour so that appropriate credit can be given to wind resources when computing the capacity/load balance to determine if the required reserve margin has been met in a given year. That capacity credit value is called the Net Dependable Capacity (NDC).

Determining Wind Capacity Factors

The NDC is applied to the nameplate capacity and is used by the expansion model to meet the 15 percent reserve margin requirement. It is calculated in a six-step process and repeated for annual, summer and winter periods for both the wind and solar resources.

TVA used actual results from its wind contracts (1500 MW in Oklahoma, Illinois, Kansas and Iowa), simulated and actual data for the in-valley sites, and proposals for various projects to determine the capacity factors for the wind resources options included in the IRP. Since each of the options originates from different regions, TVA used a region-specific estimate for annual capacity

130


Appendix B

60

Percent

50 40 30 20

Anomaly (oC)

1. For each year of the sample period, select the summer season (June-Sept). • T VA focuses this process on the summer because the system peak occurs in that season. 2. Identify the top 20 load days of the summer. • Using the top 20 days in the summer produces a distribution of wind generation in the sample year. 3. Find the peak hour for each of those top 20 days. 4. Determine the wind generation for each of those 20 peak hours and convert to capacity factors. • These generation values are converted to capacity factors by dividing the hourly generation by the nameplate capacity of the wind resource. 5. Choose the 25th percentile of this capacity factor distribution. • T VA selects the 25th percentile value to ensure that wind generation at the time of the system peak will exceed this value 75 percent of the time. 6. Then these 25th percentile annual capacity factor values are averaged across all the years of the sample to produce the NDC used for planning purposes.

10 3 2 1 0 −1 −2 −3

1980

1984

1988

1992

1996

Time

2000 2004 2008 2012 Mean capacity factor = 30.9%

Figure B‑1: Example of Wind Monthly-mean variability of net power capacity by 3TIER The Annual NDC was calculated as 14 percent based on a portfolio view of all current wind contracts to capture the diversity of location across the different states of the region. This 14 percent NDC was used for all wind options. Specific sites of future wind in MISO, SPP or in-valley is unknown, so it would be inappropriate to assume a better or worse NDC at this time. A more specific NDC would be incorporated into the wind portfolio NDC calculation once specific sites are known. TVA did not consider over-subscription contracts where transmission is limited to a level below the nameplate rating of the wind capacity which tends to improve both the annual capacity factor and the NDC rating. The costs associated with the wind projects modeled in the IRP do not reflect oversubscription; in TVA’s experience with several existing wind contracts, this over-subscription provision is negotiated in the terms and costs of a particular contract and is not easily comparable to industry benchmarks.

For this IRP study, TVA repeated this calculation using 16 years of data ranging from 1998 to 2013. The simulated hourly wind generation was provided by 3TIER, a third-party company specializing in renewable energy assessment and forecasting. The wind generation was based on simulation of TVA’s existing wind contracts in MISO, SPP, and PJM as well as a site in Kansas near where the HVDC site is proposed. 3TIER assessed the long-term variability of the wind for each site in a retrospective analysis of historical wind speed and power. These data points were derived from a mesoscale Numerical Weather Prediction (NWP) model that was statistically calibrated to match the observed data during the measurement period at the height of the towers. An example of the variability of the wind net power is shown in Figure 1.

Solar Modeling Similar to wind, solar resources are also weather and location dependent. Modeling of solar options in the IRP proceeds in a similar fashion to wind, and requires determination of solar shapes, capacity factors and NDC values. Solar data was provided by members of the TVRIX stakeholder group who commissioned

131


Appendix B

Clean Power Research (CPR) to provide TVA with the solar energy profiles for 26 sites across the Tennessee Valley shown in the map below. CPR provided SolarAnywhere® data for 15 plus years of consistent, validated, time-series irradiance measurements that provided the historical basis for the NDC, capacity factors and hourly energy patterns.

Figure B‑3: Solar Fixed Axis and Utility Tracking Capacity Factors by Month

Solar NDC values The determination of the NDC for solar resources utilizes the same process described for wind resources. The figure below shows the range of NDC values for solar fixed-axis systems computed using data covering the period 1998-2013: Figure B‑2: Sites across Tennessee Valley with historical solar irradiance data supplied by CPR

Solar Capacity Factors Using the data supplied through CPR, TVA determined that annual capacity factors are 20 percent for the fixed axis and 23 percent for the single-axis tracking option. The monthly capacity factors vary as shown in the following chart.

Figure B‑4: NDC by hour of the top 20 peak load days of Summer 1998-2013

132


Appendix B

In the summer, TVA normally has a peak load at 5:00 p.m. EST, but can also see a peak load between the hours of 2:00 p.m. and 6:00 p.m. EST. The 25th percentile of solar generation of those hours would occur at 5:00 p.m. or 6:00 p.m. EST as the sun is setting. Therefore, the summer NDC was set at 50 percent for fixed axis, including utility scale, small and large-commercial. The utility tracking option has a 68 percent NDC. All solar options have a 0 percent NDC during the winter, since TVA’s winter peaks normally occur at 5:00 a.m. EST when solar is not available.

133

Appendix C

Distributed Generation Evaluation Methodology

135


Appendix C

Background

commercial. To represent the load profiles associated with DG penetration in these customer groups, an on-site natural gas plant was assigned to industrial customers, while small solar was utilized for residential/ commercial customers. Although an assortment of DG technologies could realistically be deployed, these two technologies serve as useful proxies to represent DG across customer classes. Figure 1 shows how DG penetration, along with other uncertainties, was represented across the various scenarios and how the customer classes discussed were applied to DG penetration.

Distributed generation (DG) is broadly defined as generation that is produced on the distribution grid network. IRP strategies primarily focus on central station or utility-scale resource planning options, therefore the contributions from DG represented in this IRP are primarily captured in scenario development. In the context of the selected IRP scenarios, DG is more narrowly defined as customer-driven, demandside generation which results in utility load reductions. Additionally, DG was subdivided into two customer categories: Industrial customers and residential/

Figure C‑1: DG Market Segments and Penetration Levels across IRP Scenarios

136


Appendix C

Methodology

The primary, or leading, driver was another IRP scenario uncertainty, CO2 regulation. CO2 regulation was viewed as the most likely driving force to impact future levels of renewable energy growth, both from a utility- and customer-led perspective. Therefore, CO2 assumptions were first applied to determine utility-driven, national renewable energy adoption rates. National renewable energy adoption rates in turn drove customer-led DG renewable growth. Finally, national levels of DG growth were then appropriately scaled down to reflect regional DG growth in the Tennessee Valley region.

Different methodologies were applied to forecast DG penetration growth differences between customerled Industrial and Residential/Commercial market segments. Although the approaches differ, DG penetration levels across all scenarios directly impact other scenario uncertainties, specifically commodities, electricity prices and loads.

Residential/Commercial Distributed Generation Penetration

To begin this sequential analysis, first, CO2 uncertainty levels were correlated to traceable source data. The Reference case along with the GHG 10, GHG 15, and GHG 25 cases of the 2013 U.S. Energy Information Administration (EIA) Annual Energy Outlook were chosen as the source material. EIA and TVA CO2 price assumptions were correlated to interpolate reasonable national renewable adoption levels by 2040, EIA’s end of analysis period.

Residential/Commercial DG penetration is defined as TVA’s residential and commercial customers’ energy consumption that is self-generated by renewable energy. Renewable energy encompasses all traditional renewable resource types (solar, wind, hydro, biomass, geothermal). For the purposes of this analysis, all Residential/Commercial DG is assumed to be solar PV. To determine Residential/Commercial DG adoption rates, a sequential set of linear drivers were applied.

Figure C‑2: Correlation of IRP CO2 uncertainty values to EIA source data

137


Appendix C

The national renewable energy adoption levels, adapted from EIA data, were then adjusted to develop corresponding national DG penetration levels. EIA’s reference case and GHG 15 growth curves were

applied to national DG growth rates to ensure relative consistency between IRP scenarios. The percentages of renewable growth as a function of total renewables growth were determined as described in Figure 3.

Figure C‑3: Development of National Renewable DG Penetration Levels Figure 4 charts national renewable energy growth rates as a percentage of total generation for the electric sector (utility-led only) and including DG (customer-led). The

marginal gap between each set of solid and dashed lines indicates the quantities of DG penetration occurring across the various IRP scenarios.

Figure C‑4: National Renewable Energy Adoption Levels (Utility-led and DG)

138


Appendix C

Finally, to translate national renewable DG adoption levels to TVA regional DG levels, a 75% multiplier was applied to represent regional differences. As mentioned previously, Residential/Commercial DG penetration was assumed to be 100% solar PV to serve as a relative proxy for renewable DG growth. These growing levels

of Residential/Commercial DG penetration result in varying levels of TVA load loss as shown in Figures 5 & 6. Cumulative and annual capacity growth levels are also shown to provide a sense of total and incremental growth levels of renewable DG.

Figure C‑5: Residential/Commercial DG Adoption Levels (by 2040)

Figure C‑6: Residential/Commercial DG Adoption Levels (Annual)

139


Appendix C

Industrial Distributed Generation Penetration To accent the Residential/Commercial DG penetration analysis, industrial DG was also applied to reflect DG growth beyond renewable energy, namely from natural gas pursued by industrial customers. Industrial DG was applied across two scenarios: The Distributed Marketplace and Growth Economy. The following assumptions were applied across each scenario: Distributed Marketplace Scenario: Assumed 50% of industrial customer load was lost to DG over the study period (representing 10% of total TVA load). Growth Economy Scenario: Assumed 10% of industrial customer load with high steam needs were lost to DG over study period (0.6% of total TVA load).

Conclusion DG, both nationally and at the TVA level, is included in the 2015 IRP study as demand-side generation that is customer-driven (outside of utility involvement), and results in a reduction to utility load. Industrial DG is load loss occurring from natural gas projects while Residential/Commercial DG is represented by solar PV projects. Residential/Commercial DG is driven by CO2 regulation and national renewable and DG growth rates. The resulting combination of both Industrial and Residential/Commercial DG growth rates are captured across the various IRP scenarios as load loss. The Distributed Marketplace scenario represents the most extreme load loss on the TVA system projected to be caused by DG.

140

Appendix D

2015 IRP: Modeling Energy Efficiency

141


Appendix D

1 Energy Efficiency in the IRP

was developed in the 2015 IRP to employ “blocks” of energy efficiency impacts and costs that reflect the characteristics of existing programs but do not require the development of detailed program designs.

One of TVA’s goals is to provide low-cost, clean, and reliable electric power to consumers and it does this by maintaining a diverse set of energy resource options. Energy efficiency and demand-side management programs have been part of TVA’s energy portfolio since the late 1970s and include incentive programs, price structure changes and educational efforts to encourage awareness and smart consumer choices. TVA continues to offer programs under the EnergyRight® Solutions brand that include residential, commercial, industrial, renewable (end-use-generation), demand response and educational/outreach initiatives.

TVA energy efficiency programs typically address the major components of energy consumption in the areas of lighting, building shell improvements, HVAC/control upgrades, industrial process changes and a newly identified approach, voltage regulation. Assumptions on changes to load shapes and reductions in demand and energy can be derived from the results of existing programs and projected for blocks which serve as proxies of yet-to-be-defined future programs, as well as continuation of existing efforts. This approach greatly reduces the staff time needed to develop modeling inputs and, if designed in small enough blocks, affords the opportunity for the model to select an optimum level of energy efficiency on an annual incremental basis to match the given strategy and scenario inputs in each model run.

TVA is currently engaged in evaluating new programs, delivery and impacts as it continues to evolve the demand side management portfolio. These programs help reduce reliance on power purchases from other suppliers, reduce power production environmental impacts and mitigate utility bill pressures by providing benefits to consumers and the TVA system. Refining the characterization of energy efficiency in models will enhance potential for success and assist in keeping electricity costs low.

1.1 TVA Energy Efficiency Program Characteristics A variety of delivery methods are used to deliver programs to end-use consumers. Residential programs are delivered through various channels, which include: up-stream incentives to manufacturers and installers; promotion and administration of TVAdesigned programs through local power companies (LPCs); turnkey administration of TVA-designed programs through third-party vendors; and design, promotion and administration of programs by LPCs. In the commercial and industrial sectors, programs are offered to large customers directly served by TVA. The majority of promotion and administration duties for LPC commercial and industrial customers are handled by TVA field staff and a third-party administrator under contract to TVA with the collaboration and coordination of the LPCs. The Conservation Voltage Reduction (CVR) program requires the participation of the individual LPCs and does not involve promotion or participation by individual end-use customers.

1.0 Energy Efficiency Modeling TVA’s 2011 IRP used discrete energy efficiency portfolios matched to specific strategies for the modeling effort. The portfolios consisted of detailed program designs for individual energy efficiency and demand response programs that outlined annual costs and demand/energy reductions across a 30-year planning horizon. In the 2011 IRP, energy efficiency consisted of over 20 individual program designs, and the portfolios were considered “must run” components of their respective strategies. Two significant drawbacks to this approach were the lack of flexibility in the must run nature of the energy efficiency contribution for each strategy design and the staff time required to develop program details for efforts that would not necessarily launch for several years. To address these deficiencies, a different approach

142


Appendix D

Energy efficiency programs impact the system to reduce costs through demand reduction as well as energy savings. As can be seen in Figure 1, on a typical peak day, the energy efficiency resource provides load matching to TVA’s overall load requirements for that day. This is due to the EE resource portfolio design having

the same system load shape drivers as the system load. The variable EE shape over the majority of the day (Figure 1) and year round EE (Figure 2) demonstrates that EE resembles the cycling nature of an intermediate resource like a natural gas combined cycle unit.

Figure D‑1: Energy Efficiency Performance on a Typical Peak Summer Day (2023)

143


Appendix D

Looking across a typical year (Figure 2), energy efficiency resources provide fuel and operating cost savings by lowering demand across all months of the year and offsetting the need for base load and

intermediate resources. The shapes differ by sector with the residential sector following weather patterns more closely than the commercial or industrial sectors.

Figure D‑2: Energy Efficiency Monthly Profile (2023)

In the block designs used for the 2015 IRP, the residential sector has a defined capacity factor of 57%; the commercial sector has a capacity factor of 68%; and the industrial sector has a capacity factor of 80%. These capacity factors are comparable to other base load and intermediate duty resources with capacity factors typically greater than 40%.

The Block Concept Traditional supply side resources have the following characteristics: • Capacity and energy - typically a known size in MW and MWh respectively • Install cost - typically a bus bar $/kW • Construction lead time - years to build from initial project consideration • Operational characteristics—must run number of hours per year, heat rate (fuel efficiency), capacity factor, etc. • Service Life - years

2.0 Model Inputs and Assumptions For energy efficiency to be a selectable resource option in the optimization model, energy efficiency block characteristics must be developed that are conceptually comparable to other supply side resources.

144


Appendix D

TVA developed energy efficiency options in a similar fashion. Blocks of energy efficiency impacts and load shapes were constructed for three market sectors: residential, commercial and industrial. Each sector has a load shape similar to the weighted average of the end-

use load shapes for current EE program within those sectors. For example, a residential EE block has a load shape similar to the weighted average of six residential customer programs’ annual load shapes (Table 1). Each of the sectors is comprised of three pricing tiers

Table D‑1: Tier, Sector, and Block Hierarchy

Residential Programs

Load shapes, contribution percentages and other program characteristics of the blocks are based on the detailed Program Design Templates developed as part of the FY 2015 TVA budget. Cost and impact estimates for the blocks use an average steady-state, fully-operational estimate of program designs rather than trying to reflect the variation of higher initial/end-of-life program costs.

New Homes Self Audit In Home Energy Evaluation Manufactured Homes Heat Pump eScore Industrial Programs

Blocks were grouped by sector based on commonality of market and similarity of load shape. Each sector’s block is composed of different TVA EE programs that carry different weights. Weighting for each sector is found in Table 2 and is based on the past and projected contributions of the various programs.

Tailored Solutions for Industry Custom Industrial Standard Rebate Commercial Programs Custom Commercial Standard Rebate

Table D‑2: Weighting of EE Programs

145

Block Weight 12% 2% 20% 16% 10% 40% Block Weight 54% 10% 36% Block Weight 10% 90%


Appendix D

Each block was developed to be 10MW and between 50-72 GWh in size. This size was chosen to provide flexibility for model selection by being a proxy for EE programs. Current programs each have a net-to-gross (NTG) design assumption (Table 3) which accounts for free-ridership and other aspects of program efficacy and were weighted in the development of the sector blocks. Each existing program also has an

associated set of modeled data including the on-peak capacity reduction and associated “operational like” characteristics, which include an 8,760-hour load shape consistent with the sector end-use load shape. Since each EE block occurs at the end use level, the characteristics are “grossed up” for transmission and distribution losses to create a “supply side equivalent” when modeled with other resource options.

Program

Sub Program

Lifespan

NTG

R1

New Homes

15

64%

R2

Kit & Self Audit

6

75%

R3

IHEE

18

80%

R4

Manufactured Homes (VHP)

15

80%

R5

Heat Pump Program

15

67%

R9

ESTAR Man. Homes

15

80%

R14

eScore

18

80%

C1

Tailored Solutions

10

70%

C2

Custom Industrial

10

70%

C3

Custom Commercial

15

76%

C10

Standard Rebate Commercial

15

69%

C11

Standard Rebate Industrial

15

74%

Table D‑3: Net to Gross ratios and Lifespans for the EE programs within sectors

2.0.1 Pricing

accordance with the weighting table referenced above.

Once the operational characteristics of each sector EE block was developed, pricing tiers were identified. Pricing tiers were developed to reach more deeply into the pool of potential savings in the Valley; additional costs would need to be incurred to expand delivery system infrastructures and encourage greater participation. Blocks within Tier 1 were priced at the current portfolio of programs for each sector in

Tier 2 and Tier 3 consist of programs yet-tobe-developed (some of which represent as-yetundeveloped technologies) and pricing was based on the step function increase found in Table 4. The breakpoints and step function increase for each of the sectors were developed through consultation with the managers of existing TVA programs and supporting consultants.

146


Appendix D

Average Unweighted Increases Relative to Base Tier 2

Residential

Industrial

Commercial

ERS Incentives

50%

70%

70%

ERS Variable Costs

26%

70%

70%

ERS Fixed and Low Variable

15%

10%

10%

ERS Other

19%

70%

70%

Tier 3

Residential

Industrial

Commercial

ERS Incentives

100%

200%

200%

ERS Variable Costs

51%

200%

200%

ERS Fixed and Low Variable

25%

20%

20%

ERS Other

29%

200%

200%

Table D‑4: Tier Step Changes The steps in cost for tiers 2 and 3 are similar to a supply stack in which programs with the highest potential are the lowest cost programs, programs with mid-potential are mid cost programs, and programs with lowest to mid potential are at a high program cost. As benefits are exhausted from of the lowest cost programs, it moves down the supply stack to the next lowest cost program.

Levelized costs for each of the tiers within the sectors can be found in Figure 4. Energy efficiency programs are compared against a greenfield combined cycle plant, which energy efficiency tends to closely resemble based on capacity factor. All block costs, including incentives, escalate at inflation (1.8% per year) so that energy efficiency becomes cheaper over time in real terms.

Figure D‑3: Levelized EE block cost comparison ($/MWh) compared to a greenfield combined cycle plant over time

147


Appendix D

2.0.2 Quantity

to expand the delivery infrastructure from one year to the next and the expectation of increasing consumer/ participant awareness.

Much like the supply side counterparts, EE programs also have operational-like limits on the ramp rate, or year-over-year growth, based on startup time and development of infrastructure. The limits are driven by program development, customer awareness, market penetration, participant acquisition and many other customer and market factors.

2.0.3 Block Life For supply side resources, power contracts expire and power plants reach end of life and are retired. Similarly, energy efficiency resources have useful lifespans (e.g. light bulbs burn out, lighting systems must be upgraded and heating and cooling equipment must be replaced). For the 2015 IRP, TVA has assumed each block of energy efficiency resource can be replaced with a similar block at the available price for that sector’s block.

Through 2018, TVA has a “required” energy efficiency performance as part of a 2011 EPA settlement. These programs are embedded into the TVA annual business planning cycle and are being modeled as “must run” resources for the IRP resource selection model.

The lifespans for each of the sector blocks were developed based on program composition within each sector, current program lifespan assumptions, and measure lifespan assumptions used by industry standards.

For the selectable blocks, TVA assumed that growth of the total delivered blocks each year cannot exceed 25% when compared to the previous year for years 1-5. For years 6-15, the growth rate was limited to 20%, and then is at 15% per year thereafter. These limits were based on the ability of TVA and program partners Block Design Parameters

Final Residential

Commercial

Industrial

MW per Block

10

10

10

GWh per Block

50

59

72

Growth Rate (Yr 1-5)

25%

25%

25%

Growth Rate (Yr 6 -15)

20%

20%

20%

Growth Rate (Yr > 16)

15%

15%

15%

Max Incremental Blocks per Year Tier 1

9

4

4


7

4

2


8

4

2

Max Incremental Blocks per Year Tier Total

22

12

8

Lifespan Tier 1 (Years per Block)

17

15

12


13

13

10


13

13

10

Table D‑5: Block Characteristics for each sector

148


Appendix D

Each of the blocks in the different tiers and sectors has differing lifespans. Tier 1 block lifespans were determined using a weighted average based on existing programs. Tier 2 and 3 blocks are made up of programs yet to be developed as well as some potentially unknown technologies, therefore the same estimates could not be applied. TVA instead used an industry average lifespan for each of the sectors. Residential and commercial tier 2 and 3 blocks have a 13 year lifespan and industrial tier 2 and 3 blocks a 10year lifespan (Table 5).

There are two basic ways to incorporate the EE shapes into System Planning models: 1) As a load modifier: the energy efficiency shapes are subtracted from the original system load and the resulting net system load is fed into the model’s “load input.” 2) As a resource (selectable or non-selectable): consistent with how all other supply side resources are modeled (i.e. nuclear, coal, gas, hydro, etc.). EE resources point to a defined energy pattern (i.e. the EE load shape) similar to a solar resource.

3.0 Energy Efficiency Methodology within System Planning

Each approach has pros and cons and the best approach depends on modeling architecture and modeling objectives. For the 2015 IRP, TVA elected to use the model-as-a-selectable-resource approach. This allows TVA to model selectable EE resource units for full optimization. Energy efficiency is nondispatchable and operates similarly to a number of other non-dispatchable generation resources in that system operators cannot directly control it based on system needs. There are no variable operations and maintenance (VOM) costs nor an emissions penalty (CO2 costs). Key input parameters are monthly avoided capacity, $/kW (cost divided by summer peak kW) and an hourly energy pattern.

3.1 Planning Approach Energy Efficiency (EE) programs have two basic impacts that are relevant to planners: 1) A voided energy calculation – Energy not consumed means fuel not burned, resulting in savings in variable costs. Further, since program impacts are felt at the meter, they also avoid transmission and distribution (thermal) losses which can average 6.5% by the time energy reaches an end user. 2) A voided capacity calculation – Capacity is avoided, because reduced electricity demand translates into reduced need for incremental capacity additions.

3.2 New Approach to Modeling from 2011 IRP

Using EE program design parameters, hourly demand profiles are developed via engineering models, such as eQuest, and then calibrated through program evaluation. Inputs to the models include occupancy/ utilization profiles and weather data. Each model’s key output is an 8,760 hourly profile of a “before” end use shape and an “after” efficient end use shape that are subtracted to get the net savings. The net savings shape is then regressed on weather and calendar variables, revealing the relationship between savings and temperature, day of week, season, etc. The model is then forecast forward using TVA weather and load forecast as inputs. The final result is an hourly energy efficiency savings forecast synched to the TVA load forecast.

TVA is taking a new approach to energy efficiency modeling to allow energy efficiency to compete with other resources within each of the IRP cases. This will create an opportunity to allow for full portfolio optimization, to better gauge the impacts of the programs in different situations, and to better demonstrate the value proposition for the resource. The 2011 IRP study did not contain energy efficiency as a selectable resource. Several different EE portfolios were scheduled as load modifiers in various scenarios. There was no supply stack concept in those portfolios, which in effect reduces model flexibility and limits model outcome. TVA’s new modeling approach for

149


Appendix D

3.2.1 Comparability to Other Supply Side Resources

energy efficiency as a competitive resource attempts to enhance model visibility and potential impacts with regards to least-cost optimization.

Energy efficiency unit characteristics must be developed that are comparable to other supply side resources. Supply side characteristics that feed the capacity expansion model can be found in Table 6 and are compared against the energy efficiency “power plant.” SUPPLY SIDE COMPARISON

Year Available

Comm EE

Ind EE

Res EE

New CC

New CT

2014

2014

2014

2019

2018

Outage Rate Heat Rate Fuel Costs Fuel CAGR CO2 Costs CO2 CAGR (starts in 2022) O&M costs O&M Escalation Transmission Contingency Cost Project Contingency Cost Capital Costs Escalation of Capital Capacity Factor Technology Shifts

Table D‑6: Resource Characteristic Comparison with EE

150

New Coal AP 1000 w/ CCS 2028

2026


Appendix D

For supply side resources in the IRP, unit performance is not expected to be 100%. This delivery risk is captured in an outage rate for the unit. There is not a comparable outage rate for the modeled energy efficiency blocks; rather, the modeling approach assumes the block to be operationally available 100% of the time. Efficiency is dependent on variables such as equipment reliability and service life, operating conditions, etc., that would impact operability similar to an outage rate.

combined cycle plant constructed in 2033 possesses the same heat rates, ramp rates, cost of construction (escalated for inflation), etc. as one constructed in 2015 because we do not know what the future technology will be. However, in EE blocks TVA allows for an assumption of technological improvements based on the history of EE deliveries over the past 30 years.

3.3 Modeling Uncertainty The block design approach is novel and fits well with model architecture, but introduces some uncertainties around design and delivery that are unique relative to other resources. Design uncertainty is introduced by the creation of prescribed blocks of EE meant to reflect bundles of programs over time. Delivery uncertainty exists around claimed versus evaluated measures, the ability to deliver and implement programs though TVA’s 155 different local power companies, and risk around EE deliveries relative to future codes and standards.

In addition to outage rates, Table 7 shows the potential uncertainties that are captured in cost for supply side resources. Examples include a carbon dioxide emission penalty, fuel cost uncertainty, project cost contingencies and cost escalation uncertainties. One item unique to TVA’s modeling approach on EE blocks is related to technological improvements. Traditional supply side resources do not reflect advancements in technology over time. For example, a

Uncertainty Design

Deliver

Proxy Programs in Blocks

LPC Delivery Risk

Measure Lifespand Blending

Codes and Standards

Unchanging Shapes

Claimed vs. Evaluated

Table D‑7: Design and Delivery Uncertainties

3.3.1 Design Uncertainty

Uncertainty of all types exists with supply side resources and is modeled in different ways in the analysis, but typically manifest itself as cost. For energy efficiency, TVA considers the two primary categories of uncertainty mentioned above to remain comparable with other supply side resources. In addition, certain variables can be captured directly or indirectly in the stochastic analysis performed in the study. Key uncertainties are discussed in more detail below.

Since the modeled energy efficiency blocks are proxies for technologies and programs not yet developed, there is uncertainty in their design and future composition. Blocks in the study are modeled as 10 MW resources with a defined load shape by sector (residential, commercial, and industrial). The virtual nature of energy efficiency compared with the tangible, physical attributes of supply side resources necessarily introduces a level of uncertainty around certain key

151


Appendix D

design attributes.

for overestimation or underestimation of energy savings. With respect to the energy shapes, the capacity expansion model uses a repeating annual energy pattern for each block to the end of the lifespan. As programs die off before the expected lifespan, they are replaced with the same technology at no cost until the end of a defined block life.

3.3.1.1 Measure Life Uncertainty Measure life or Effective Useful Life (EUL) is the median number of years that the measure after installation is expected to be in place and operable. This includes “equipment life” which is the number of years installed equipment will be operational before it fails, and “measure persistence” which takes into account business turnover, failure or early retirement of the installed equipment.

Figure 5 demonstrates how this applies to a 14-year residential audit program within a residential block that has a lifespan of 17 years. Several technologies die off before the end of life, but the block assumes the energy is still there because the technology is replaced with like kind (solid black line). Notice there is an overstatement of energy for years 6 -17. For the technologies where contribution ends prior to end of block life, it is replaced with a similar block and contributes with the same energy pattern for the remaining block life. The risk in these cases is that we are overstating energy (by having the same energy contribution every year) and underestimating costs (by assuming technology is replaced at no cost to TVA). The blending of programs into blocks creates unique challenges for resource

Each of the energy efficiency blocks contains different programs with different EULs. Tier 1 blocks contain currently developed TVA programs, and the block lifespan was determined using weighted averages. Block lifespans for tiers 2 and 3 were approached differently. Because tiers 2 and 3 contain undeveloped technologies and programs, industry average standards were used for the different sector’s lifespan. Since the energy efficiency blocks are a mix of differing technologies with differing life measures, potential exists

planning in that an average lifespan can create resource adequacy challenges in a particular year.

Figure D‑4: Example of a residential audit program modeled as part of a residential block

152


Appendix D

accounts for attribution) in order to get “the net realization rate.” In most studies reviewed by TVA, net realization rates tend to be less than one, although in some jurisdictions realization rates reflecting actual performance exceeding planned savings have been achieved.

3.3.1.2 Fixed Shape Uncertainty Each of the energy efficiency sectors has a fixed end use 8,760 load shape. For modeling convenience, the blocks are assumed to have an unchanging composition over time, even though this is unlikely. TVA’s stochastic modeling around the overall TVA load shape partially addresses some of this risk but does not address the uncertainty around the shape of the block designs changing over time as programs and technologies evolve and as the “low hanging fruit” of EE is picked off.

Examples of lower realization rates (i.e. realized program impacts) can be seen in more mature markets such as California, Con Edison and Indiana where there are extensive measurement and verification (M&V) data. They illustrate that risk exists with regard to energy and capacity impacts even in these more mature markets. A lot of this is attributable to operational issues, calculation methods, and inappropriate baselines. TVA does not expect to repeat industry experience with regards to claimed and evaluated measured discrepancies since TVA has different market drivers. However, TVA can learn from their experience by noting that there is risk around these future program assumptions. In the IRP case, the risk is primarily around our ability to realize deliveries over the 20-year study period on programs that as-yet have not been designed, undergone M&V and been refined. This uncertainty increases over time.

3.3.2 Delivery Risk Uncertainty 3.3.2.1 Local Power Company Delivery Uncertainty Unlike conventional assets that can be constructed, operated and maintained directly by TVA, there is more uncertainty around the ability to implement EE resources in the Tennessee Valley because of the multiple parties involved and coordination around end use customer adoption. The end use providers are made up of the participants and the local power companies. There are currently 155 Local Power Companies (LPCs) in the TVA region consisting of municipal utility companies and cooperatives. Since TVA is not the end-use provider there is risk in how the 155 local power companies would vary in their delivery of EE programs. Additionally, TVA and the LPCs need to establish delivery mechanisms to facilitate larger EE deployment across the region and this takes time and resources which may be different than a comparable, vertically integrated utility might experience.

3.3.2.3 Delivery Risk: Codes and Standards TVA’s modeling approach assumes that selectable EE resource deliveries are over and above any future tightening of efficiency codes and standards. Currently, known codes and standards (C&S) are reflected in the load forecast in the IRP, and future increases in C&S are not assumed. Treating EE as a supply side resource means that it is available and deliverable in the same way that a conventional resource is, and this creates a risk around C&S tightening. A conventional gas turbine for example, delivers MWs regardless of whether new efficiency standards reduce TVA’s sales in year 15 of the study. For EE, there is a risk that future tightening of C&S would reduce the amount of EE available to deploy in the market or increase the cost of deploying the EE resource in the future. As baseline efficiency requirements increase, then either the supply (volume) of EE must decrease or the cost of the next series of

TVA believes delivery risk will diminish over time as delivery mechanisms are developed and refined with the LPC customers. A 10% adjustment is applied to reflect delivery risk for years 1 through 5. At year 6, this adjustment begins declining at 2% per year.

3.3.2.2 Realization Rate Delivery Uncertainty The gross realization rate is the ratio of measured energy reduction (actual) to claimed energy reduction (planned). The gross realization rate is typically multiplied by the Net to Gross ratio (a ratio which

153


Appendix D

measures must increase. TVA’s current EE modeling assumes that over the 20-year study period that TVA programs can be developed to exceed whatever the then-current standards may be.

our service territory and build the infrastructure with our LPC partners. This is represented as a 10% cost adder in years 1-5 that begins to decline in year 6. The other uncertainties around block design and delivery risk uncertainty are initially zero but begin to grow over time, starting in year 6. The total planning adjustment is shown in Figure 6 and grows to 30 percent over the out years of the study. This planning adjustment reflects the fact that the further out in the future one goes, the more uncertain these proxy EE blocks are. The planning adjustment is an approximation, not a precise calculation, but is meant to reflect how uncertainties increase over time.

3.4 Recognizing Design & Delivery Uncertainty: Planning Factor Adjustment Why do all these performance issues and uncertainty matter? Dynamically modeling energy efficiency as a resource means that all variables, including resource costs, shapes and uncertainties, significantly influence the modeled needs for base load, intermediate and peaking generation. There are several possible ways to address this uncertainty analytically, including carrying higher planning reserves, but each increases overall plan costs. To address these uncertainties and allow energy efficiency to compete on the same playing field as a supply side resource, a planning adjustment factor was made to reflect the two categories of design and delivery uncertainty.

In this construct, the uncertainties manifest as cost in the model. The alternate approach was to restrict volumes available in the out years, but TVA chose to keep the volumes consistent to test the model boundaries. Uncertainty manifesting as cost has certain modeling advantages and also allows volumes to be unconstrained. In many case results we can see full selection of EE blocks occur, even in the out years with the uncertainty adjustments, which allows for a more robust range of case results.

Initially, the primary risk TVA faces is delivery risk, largely around the ability to implement programs across

Figure D‑5: Planning Factor Adjustment over time 154


Appendix D

3.5 Recognizing Uncertainty: Stochastic Analysis

varied by demand and weather pattern (i.e. load shape) distributions modeled in the analysis. Traditional supply side resources have other factors that can change both their cost and generation levels: demand, fuel, O&M, capital costs, CO2 emission penalties, etc. All such uncertainties manifest as cost in the model. Table 8 lists the direct and indirect stochastic variables for several supply side resources as a comparison to energy efficiency.

While the planning adjustment captures design and delivery uncertainty, TVA’s analytical approach also considers stochastic analysis on several key inputs. O&M cost escalations are stochastically varied in the analysis using the same distributions as other O&M costs. Resulting system cost impacts are indirectly

direct indirect

Stochastic Variables Diesels

CT

CC

Coal

Nuclear

Hydro

Solar

Wind

Energy Efficiency

Gas price Coal price Oil price CO2 allowance price Electricity price Hydro generation Plant availability Load shape year Electricity demand O&M costs Interest rates Capital cost

Table D‑8: Indirect and Direct Stochastic Variables

Even after accounting for the planning factor uncertainty, EE blocks have a significantly lower range of uncertainty than a comparable combined cycle plant as shown in Figure 7. The uncertainty bands around combined cycle costs are much wider due to fuel, emissions, O&M, capacity factor and capital

cost uncertainty. The much narrower EE uncertainty band is driven by the design and delivery uncertainties previously covered, stochastic variations on O&M cost and the indirect effects of the stochastic draws on the overall system load shape.

155


Appendix D

Figure D‑6: Uncertainty bands in $/MWh for each of the EE sector blocks as compared to a greenfield combined cycle plant

156


Appendix D

3.6 Costs after Planning Adjustment

Looking at the LCOE over time with the uncertainty adjustment, most of the EE blocks remain less expensive than a natural gas combined cycle unit through the IRP study period. Only Residential Tier 3 has block costs that are higher in the beginning and end of the study period than a comparable combined cycle.

Levelized Cost of Energy (LCOE) is a common metric to allow comparisons of total resource costs reflective of capital costs, asset lives and expected fuel costs. Looking at the comparison in Figure 8, EE compares favorably with other TVA resources in 2015.

Figure D‑7: Levelized Cost comparisons in 2015 (2015$/ MWh) Figure D‑8: Levelized cost comparison ($/MWh) of EE tiers in 2015 and 2033

Figure D‑9: Levelized Cost Comparison by Sector through time

157


Appendix D

4.0 Next Steps Energy efficiency modeling for the IRP was a collaborative effort across TVA and with stakeholders. Modeling energy efficiency as a competitive resource introduces additional uncertainties around design and delivery that are unique from other traditional resources. TVA’s approach accounts for these uncertainties with a planning adjustment, which is hoped to refine over time as programs are developed, measured and verified. The modeling framework chosen for use in the 2015 IRP has produced a robust set of results that demonstrate the value energy efficiency brings to the portfolio, including an assessment of the outcome for cases that test the boundaries for EE. TVA’s next step is to develop an internal business process to leverage this dynamic approach in resource planning and to revisit the assumptions behind some of the fundamental parameters.

158

Appendix E

Capacity Plan Summary Charts

159


Appendix E

Capacity & Energy Expansion Results Appendix

over the planning horizon. The capacity is in gigawatts, which is 1,000 megawatts, and is based on the summer net dependable capacity value or the amount of capacity that TVA plans to have available to meet summer peak firm requirements.

The capacity expansion plans are shown below by strategy. The capacity graphics show the total capacity grouped by resource type (i.e., nuclear, hydro, coal, etc.) Total Capacity Expansion Plans

Strategy A: The Reference Plan GW, SND 50



Scenario 3: Economic Growth



45 40 35 30 25 20 15 10 5 0 DR EE Gas Renewables Coal Hydro Nuclear

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 10 13 13 15 16 0 1 1 2 3 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 3 10 12 12 13 15 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 11 14 14 15 16 0 1 2 3 4 11 8 8 8 8 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 11 11 12 0 1 2 3 4 11 8 7 5 5 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 3 10 12 11 12 13 0 1 1 1 2 11 8 7 5 5 4 4 4 4 4 7 8 8 8 8

Strategy B: Meet an Emission Target GW, SND 50







2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 10 13 13 15 16 0 1 1 2 3 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 3 10 12 12 13 15 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 11 13 14 15 16 0 1 2 3 4 11 8 8 8 8 5 5 5 5 5 7 8 8 8 8

160

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 11 11 13 0 1 2 3 4 11 8 7 5 5 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 3 10 12 11 12 13 0 1 1 1 2 11 8 7 5 5 4 4 4 4 4 7 8 8 8 8


Appendix E

Strategy C: Market Supplied Resources GW, SND 50







2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 10 13 12 13 15 0 1 1 2 3 11 8 8 8 8 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 3 10 12 11 12 13 0 1 1 2 2 11 8 8 8 8 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 11 13 13 13 16 0 1 2 3 4 11 8 8 8 8 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 11 11 12 0 1 2 3 4 11 8 7 6 5 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 1 2 0 1 2 3 3 10 12 11 11 11 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

Draft Strategy C: Market Supplied Resources GW, SND 50







2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 10 13 13 16 10 0 1 1 2 3 11 8 8 8 8 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 12 13 15 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 3 3 11 13 13 14 15 0 1 2 3 4 11 8 8 8 8 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 11 11 12 0 1 2 3 4 11 8 7 6 5 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 10 12 11 11 11 0 1 1 2 2 11 8 8 7 6 4 4 4 4 4 7 8 8 8 8

*The original construct of the draft strategy C contained shorter term power purchase agreements. The revised strategy C only included 20-year commitments. See section 7.1.2 for further information.

161


Appendix E

Strategy D: Maximize EE GW, SND 50







2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 4 5 10 13 13 14 15 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 4 5 10 12 12 12 13 0 1 1 2 2 11 8 8 7 7 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 4 5 11 14 14 15 15 0 1 2 3 3 11 8 8 7 7 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 4 5 10 12 11 11 12 0 1 2 3 3 11 8 6 5 4 4 4 4 4 4 7 8 8 8 8

2015 2020 2025 2030 2033 2 2 2 2 2 0 1 2 4 5 10 12 11 11 12 0 1 1 1 1 11 8 7 5 5 4 4 4 4 4 7 8 8 8 8

Strategy E: Maximize Renewables GW, SND 50






45 40 35 30 25 20 15 10 5 0 DR EE Gas CC Gas CT Renewables Coal Hydro Nuclear

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 5 7 6 6 6 5 6 6 6 7 0 1 3 4 5 11 8 8 7 7 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 2 2 2 0 1 2 3 2 5 7 6 6 6 5 6 6 6 7 0 1 3 4 5 11 8 7 6 5 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 2 1 2 2 2 0 1 2 3 3 5 7 6 6 6 5 6 7 8 9 0 2 3 5 6 11 8 8 7 7 5 5 5 5 5 7 8 8 8 8

162

2015 2020 2025 2030 2033 1 1 1 2 2 0 1 2 3 3 5 7 6 6 6 5 6 6 6 6 0 1 3 4 5 11 8 6 5 4 5 5 5 5 5 7 8 8 8 8

2015 2020 2025 2030 2033 1 1 1 1 2 0 1 2 2 2 5 7 6 6 6 5 6 6 6 6 0 1 2 4 4 11 7 6 5 4 5 5 5 5 5 7 8 8 8 8


Appendix E

Below are the total energy charts that correspond to the capacity expansion plans above. The energy charts show total energy grouped by resource type (i.e.,

nuclear, hydro, coal, etc.) over the planning horizon and are in terawatt hours, which is a 1,000 gigawatt hours.

Total Energy Plans Strategy A: The Reference Plan TWh 200






150

100

50

0 EEDR Gas Renewables Coal Hydro Nuclear

2015 2020 2025 2030 2033 1 5 12 17 17 23 31 30 36 37 6 6 8 10 16 58 41 39 37 34 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 17 23 26 24 30 38 6 6 6 8 6 57 40 39 37 36 16 16 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 18 25 25 28 31 29 6 7 10 14 25 58 51 45 44 39 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 18 21 13 12 17 18 6 22 25 27 32 59 30 27 23 20 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 16 16 20 22 17 19 25 6 6 6 8 5 57 36 31 27 25 16 16 16 17 16 55 67 69 67 68

Strategy B: Meet an Emission Target TWh 200






150

100

50


2015 2020 2025 2030 2033 1 5 12 17 17 23 31 30 36 38 6 6 8 10 15 58 41 39 37 34 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 17 23 26 24 30 38 6 6 6 8 5 57 40 39 37 36 16 16 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 18 25 24 28 33 27 6 7 10 12 28 58 51 45 45 38 17 18 18 18 18 55 67 69 67 68

163

2015 2020 2025 2030 2033 1 5 12 18 18 21 14 13 17 19 6 21 24 27 31 59 30 27 23 20 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 16 16 20 22 17 19 25 6 6 6 8 5 57 36 31 27 25 16 16 16 17 16 55 67 69 67 68


Appendix E

Strategy C: Market Supplied Resources Scenario 1: Current Outlook

TWh 200



2015 2020 2025 2030 2033 1 5 12 17 17 23 26 23 25 32 6 6 7 9 7 57 40 39 40 40 16 16 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 6 14 19 19 25 23 26 29 28 6 7 10 14 25 58 51 45 44 39 17 18 18 18 18 55 67 69 67 68



150

100

50

0

2015 2020 2025 2030 2033 1 5 12 17 18 23 30 29 32 36 6 6 9 11 10 58 41 39 40 40 16 17 17 17 17 55 67 69 67 68

EEDR Gas Renewables Coal Hydro Nuclear

2015 2020 2025 2030 2033 1 5 13 18 19 21 14 12 16 17 6 22 24 26 31 59 30 27 24 21 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 11 17 17 20 22 16 16 21 6 6 6 8 6 57 36 33 29 29 16 16 16 17 16 55 67 69 67 68

Draft Strategy C: Market Supplied Resources TWh 200






2015 2020 2025 2030 2033 1 5 12 17 17 23 26 24 28 36 6 6 7 9 6 57 40 39 37 37 16 16 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 6 14 19 19 25 23 26 28 27 6 8 10 15 25 58 51 45 44 39 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 13 18 19 21 14 12 16 17 6 22 24 27 31 59 30 27 23 21 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 16 16 20 22 16 17 21 6 6 6 8 6 57 36 33 29 28 16 16 16 17 16 55 67 69 67 68

150

100

50


2015 2020 2025 2030 2033 1 5 12 17 17 23 31 29 32 35 6 6 8 10 12 58 41 40 40 40 16 17 17 17 17 55 67 69 67 68

*The original construct of the draft strategy C contained shorter term power purchase agreements. The revised strategy C only included 20-year commitments. See section 7.1.2 for further information.

164


Appendix E

Strategy D: Maximize EE TWh 200






150

100

50


2015 2020 2025 2030 2033 1 5 13 24 29 23 30 29 32 36 6 6 7 9 6 58 42 39 35 34 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 13 24 29 23 26 23 24 29 6 6 6 8 6 57 40 38 35 33 16 16 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 13 24 29 25 25 28 32 25 6 6 9 11 25 58 51 45 39 33 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 13 24 29 21 13 14 15 18 6 22 25 27 25 59 30 23 19 17 16 17 17 17 17 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 13 24 29 20 22 16 15 17 6 6 6 6 3 57 36 31 26 23 16 16 16 16 16 55 67 69 67 68

Strategy E: Maximize EE TWh 200






150

100

50


2015 2020 2025 2030 2033 1 5 12 17 17 23 21 17 18 19 6 18 24 33 38 57 37 35 30 28 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 16 16 23 19 14 15 17 6 16 22 30 36 56 37 32 29 26 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 12 17 17 25 16 17 20 22 6 19 26 34 40 57 47 41 35 32 17 18 18 18 18 55 67 69 67 68

165

2015 2020 2025 2030 2033 1 5 12 17 18 21 14 14 18 18 6 21 25 29 34 58 30 23 19 17 17 18 18 18 18 55 67 69 67 68

2015 2020 2025 2030 2033 1 5 11 13 12 20 17 9 9 11 6 16 20 25 29 57 31 25 22 19 17 18 18 18 17 55 67 69 67 68


Appendix E

Below are the total capacity additions on a year by year basis. The data is shown in summer net dependable

gigawatts and is grouped by resource type (i.e., nuclear, hydro, coal, etc.) over the planning horizon.

1A, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

1B, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

1C, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

1D, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

1E, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.1 11.7 0.3 10.3 0.1 1.5 34.6

6.7 4.1 11.7 0.3 10.3 0.1 1.5 34.6 6.7 4.1 11.7 0.3 10.3 0.1 1.5 34.6

6.7 4.1 11.7 0.3 10.3 0.1 1.5 34.6

6.7 4.1 11.7 0.3 10.3 0.1 1.5 34.6

6.7 4.2 11.3 0.3 10.5 0.1 1.6 34.8

6.7 4.2 11.3 0.3 10.5 0.1 1.6 34.8 6.7 4.2 11.3 0.3 10.5 0.1 1.6 34.8

6.7 4.2 11.3 0.3 10.5 0.1 1.6 34.8

6.7 4.6 11.3 0.3 10.2 0.1 1.5 34.8

7.9 4.3 10.3 0.4 10.5 0.2 1.6 35.1

7.9 4.3 10.3 0.4 10.5 0.2 1.6 35.1 7.9 4.3 10.3 0.4 10.4 0.2 1.7 35.1

7.9 4.3 10.3 0.4 10.4 0.2 1.7 35.1

7.9 4.6 10.3 0.4 10.2 0.2 1.5 35.1

7.9 4.3 9.1 0.4 12.0 0.3 1.6 35.5

7.9 4.3 9.1 0.4 12.0 0.3 1.6 35.5 7.9 4.3 9.1 0.4 11.9 0.3 1.7 35.5

7.9 4.3 9.1 0.4 11.8 0.3 1.8 35.5

7.9 4.6 9.1 0.4 11.8 0.3 1.5 35.5

8.0 4.3 9.1 0.5 12.0 0.4 1.6 35.9

8.0 4.3 9.1 0.5 12.0 0.4 1.6 35.9 8.0 4.3 9.1 0.5 11.9 0.4 1.7 35.9

8.0 4.3 9.1 0.5 11.7 0.4 1.9 35.9

8.0 4.6 9.1 0.7 11.6 0.4 1.5 35.9

8.3 4.3 8.4 0.5 12.9 0.6 1.6 36.5

8.3 4.3 8.4 0.5 12.9 0.6 1.6 36.5 8.3 4.3 8.4 0.5 12.9 0.6 1.6 36.5

8.3 4.3 8.4 0.5 12.2 0.6 1.9 36.1

8.3 4.6 8.4 1.1 12.2 0.6 1.4 36.5

8.3 4.3 8.0 0.5 12.9 0.7 1.6 36.3

8.3 4.3 8.0 0.5 12.9 0.7 1.6 36.3 8.3 4.3 8.0 0.5 12.9 0.7 1.6 36.3

8.3 4.3 8.0 0.5 12.6 0.7 1.9 36.3

8.3 4.6 8.0 1.4 12.2 0.7 1.3 36.5

8.3 4.3 8.0 0.5 12.9 0.9 1.7 36.7

8.3 4.3 8.0 0.5 12.9 0.9 1.7 36.7 8.3 4.3 8.0 0.6 12.9 0.9 1.6 36.7

8.3 4.3 8.0 0.5 12.7 0.9 1.9 36.7

8.3 4.6 8.0 1.6 12.2 0.9 1.3 37.0

8.3 4.3 8.0 0.5 13.1 1.1 1.8 37.1

8.3 4.3 8.0 0.5 13.1 1.1 1.8 37.1 8.3 4.3 8.0 0.8 12.9 1.1 1.7 37.1

8.3 4.3 8.0 0.5 13.0 1.1 1.9 37.1

8.3 4.6 8.0 1.9 12.2 1.1 1.3 37.4

8.3 4.4 8.0 0.7 13.0 1.3 1.9 37.6

8.3 4.4 8.0 0.6 13.1 1.3 1.9 37.6 8.3 4.4 8.0 1.1 12.1 1.3 1.8 36.8

8.3 4.4 8.0 0.5 13.2 1.4 1.9 37.6

8.3 4.7 8.0 2.2 11.8 1.3 1.4 37.6

166

8.3 4.4 8.0 0.9 12.9 1.6 1.9 38.0

8.3 4.4 8.0 0.8 12.9 1.6 1.9 37.9 8.3 4.4 8.0 1.3 11.8 1.6 1.9 37.1

8.3 4.4 8.0 0.6 13.0 1.7 1.9 37.9

8.3 4.7 8.0 2.5 11.5 1.6 1.4 37.9

8.3 4.4 8.0 1.1 13.0 1.8 1.9 38.4

8.3 4.4 8.0 1.1 13.1 1.8 1.9 38.4 8.3 4.4 8.0 1.5 11.9 1.8 1.9 37.7

8.3 4.4 8.0 0.8 13.0 2.1 1.9 38.5

8.3 4.7 8.0 2.8 11.5 1.8 1.4 38.4

8.3 4.4 7.0 1.4 13.9 2.0 1.9 38.9

8.3 4.4 7.0 1.3 14.0 2.0 1.9 38.9 8.3 4.4 7.9 1.7 12.0 2.0 1.9 38.2

8.3 4.4 7.0 1.0 13.9 2.4 1.9 38.9

8.3 4.7 7.0 3.2 12.3 2.0 1.4 38.9

8.3 4.4 7.0 1.6 14.0 2.2 1.9 39.4

8.3 4.4 7.0 1.5 14.1 2.2 1.9 39.4 8.3 4.4 7.9 1.9 12.0 2.2 1.9 38.6

8.3 4.4 7.0 1.2 13.8 2.8 1.9 39.4

8.3 4.7 7.0 3.5 12.2 2.2 1.4 39.4

8.3 4.4 7.0 1.8 14.2 2.3 1.9 39.9

8.3 4.4 7.0 1.7 14.6 2.3 1.9 40.2 8.3 4.4 7.9 2.1 12.2 2.3 1.9 39.2

8.3 4.4 7.0 1.4 13.8 3.1 1.9 39.9

8.3 4.8 7.0 3.8 12.2 2.3 1.5 39.9

8.3 4.4 7.0 2.0 14.6 2.5 1.9 40.6

8.3 4.4 7.0 1.9 14.6 2.5 1.9 40.6 8.3 4.4 7.9 2.3 12.5 2.5 1.9 39.8

8.3 4.4 7.0 1.6 13.9 3.4 1.9 40.6

8.3 4.8 7.0 4.1 12.3 2.5 1.6 40.6

8.3 4.4 7.0 2.1 14.6 2.7 1.9 40.9

8.3 4.4 7.0 2.1 14.6 2.7 1.9 40.9 8.3 4.4 7.9 2.5 12.5 2.7 1.9 40.2

8.3 4.4 7.0 1.8 13.8 3.8 1.9 41.1

8.3 4.8 7.0 4.4 12.2 2.7 1.6 41.0

8.3 4.4 7.0 2.3 14.8 2.7 1.9 41.5

8.3 4.4 7.0 2.2 14.9 2.7 1.9 41.5 8.3 4.4 7.9 2.7 12.8 2.7 1.9 40.7

8.3 4.4 7.0 2.0 13.8 4.2 1.9 41.6

8.3 4.8 7.0 4.8 12.2 2.7 1.7 41.5

8.3 4.4 6.6 2.4 15.5 2.8 1.9 41.9

8.3 4.4 6.6 2.3 15.6 2.8 1.9 41.9 8.3 4.4 7.5 2.8 13.5 2.8 1.9 41.2

8.3 4.4 6.6 2.1 14.2 4.4 1.9 42.0

8.3 4.8 6.6 5.0 12.7 2.7 1.8 41.9

8.3 4.4 6.6 2.8 15.9 2.7 1.9 42.6

8.3 4.4 6.6 2.7 16.1 2.7 1.9 42.8 8.3 4.4 7.5 3.0 14.7 2.8 1.9 42.6

8.3 4.4 6.6 2.2 14.7 4.6 1.9 42.7

8.3 4.8 6.6 5.4 13.1 2.7 1.8 42.6


Appendix E


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.1 0.1 1.5 34.4

6.7 4.2 11.3 0.3 10.3 0.1 1.5 34.5

7.9 4.2 10.3 0.4 10.2 0.2 1.5 34.7

7.9 4.2 9.1 0.4 11.5 0.3 1.5 34.9

8.0 4.3 9.1 0.5 11.4 0.4 1.5 35.0

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 8.0 0.5 12.2 0.9 1.3 35.5

8.3 4.3 8.0 0.5 12.2 1.1 1.4 35.7

8.3 4.3 8.0 0.5 12.3 1.3 1.4 36.0

8.3 4.3 8.0 0.5 12.2 1.6 1.4 36.3

8.3 4.4 8.0 0.5 12.2 1.8 1.5 36.7

8.3 4.4 7.0 0.7 13.1 2.0 1.6 37.1

8.3 4.4 7.0 1.0 13.0 2.2 1.7 37.6

8.3 4.4 7.0 1.1 13.1 2.3 1.8 38.0

8.3 4.4 7.0 1.3 13.2 2.5 1.9 38.7

8.3 4.4 7.0 1.5 13.2 2.7 1.9 39.1

8.3 4.4 7.0 1.7 13.6 2.7 1.9 39.5

8.3 4.4 6.6 1.8 14.4 2.7 1.8 39.9

8.3 4.4 6.6 1.9 15.1 2.6 1.7 40.6


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.2 0.1 1.5 34.5

6.7 4.2 11.3 0.3 10.3 0.1 1.5 34.5

7.9 4.2 10.3 0.4 10.2 0.2 1.5 34.7

7.9 4.2 9.1 0.4 11.5 0.3 1.5 34.9

8.0 4.3 9.1 0.5 11.4 0.4 1.5 35.1

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 8.0 0.5 12.2 0.9 1.3 35.5

8.3 4.3 8.0 0.5 12.2 1.1 1.4 35.7

8.3 4.3 8.0 0.5 12.3 1.3 1.4 36.0

8.3 4.3 8.0 0.5 12.2 1.6 1.4 36.3

8.3 4.4 8.0 0.6 12.2 1.8 1.6 36.8

8.3 4.4 7.0 0.8 13.1 2.0 1.7 37.2

8.3 4.4 7.0 1.0 13.0 2.2 1.7 37.7

8.3 4.4 7.0 1.2 13.1 2.3 1.8 38.1

8.3 4.4 7.0 1.4 13.3 2.5 1.9 38.8

8.3 4.4 7.0 1.6 13.4 2.7 1.8 39.2

8.3 4.4 7.0 1.8 13.8 2.7 1.7 39.7

8.3 4.4 6.6 1.9 14.6 2.7 1.6 40.1

8.3 4.4 6.6 2.0 15.3 2.7 1.6 40.9


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.1 0.1 1.5 34.4

6.7 4.2 11.3 0.3 10.3 0.1 1.5 34.5

7.9 4.2 10.3 0.4 10.2 0.2 1.5 34.7

7.9 4.2 9.1 0.4 11.5 0.3 1.5 34.9

8.0 4.3 9.1 0.5 11.4 0.4 1.5 35.0

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 8.0 0.5 12.2 0.9 1.5 35.6

8.3 4.3 8.0 0.5 12.0 1.1 1.6 35.7

8.3 4.3 8.0 0.5 11.9 1.3 1.7 36.0

8.3 4.3 8.0 0.7 11.5 1.6 1.8 36.1

8.3 4.4 8.0 0.9 11.5 1.8 1.9 36.7

8.3 4.4 7.9 1.1 11.5 2.0 1.9 37.1

8.3 4.4 7.9 1.3 11.5 2.2 1.9 37.5

8.3 4.4 7.9 1.5 11.7 2.3 1.9 38.0

8.3 4.4 7.9 1.7 12.0 2.5 1.9 38.7

8.3 4.4 7.9 1.9 12.0 2.7 1.9 39.0

8.3 4.4 7.9 2.1 12.2 2.7 1.9 39.5

8.3 4.4 7.5 2.2 13.0 2.7 1.9 40.0

8.3 4.4 7.5 2.3 13.5 2.7 1.9 40.6


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.1 0.1 1.4 34.4

6.7 4.2 11.3 0.3 10.2 0.1 1.5 34.5

7.9 4.2 10.3 0.4 10.2 0.2 1.6 34.9

7.9 4.2 9.1 0.4 11.2 0.3 1.7 34.9

8.0 4.3 9.1 0.5 11.2 0.4 1.8 35.3

8.3 4.2 8.4 0.5 12.2 0.6 1.8 36.0

8.3 4.3 8.0 0.5 12.2 0.7 1.8 35.8

8.3 4.3 8.0 0.5 12.2 0.9 1.8 36.0

8.3 4.3 8.0 0.5 12.2 1.1 1.8 36.2

8.3 4.3 8.0 0.5 11.7 1.4 1.8 36.0

8.3 4.3 8.0 0.5 11.5 1.7 1.8 36.2

8.3 4.4 8.0 0.5 11.7 2.1 1.8 36.7

8.3 4.4 7.0 0.7 12.5 2.4 1.8 37.1

8.3 4.4 7.0 0.9 12.3 2.8 1.8 37.6

8.3 4.4 7.0 1.1 12.4 3.1 1.8 38.1

8.3 4.4 7.0 1.3 12.5 3.4 1.8 38.7

8.3 4.4 7.0 1.5 12.2 3.8 1.8 39.1

8.3 4.4 7.0 1.7 12.2 4.2 1.8 39.6

8.3 4.4 6.6 1.9 12.6 4.4 1.8 40.0

8.3 4.4 6.6 2.0 13.0 4.6 1.8 40.7


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.2 0.1 1.3 34.4

6.7 4.6 11.3 0.3 10.1 0.1 1.3 34.5

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.4 11.2 0.3 1.3 34.9

8.0 4.6 9.1 0.6 11.2 0.4 1.3 35.3

8.3 4.6 8.4 0.9 12.2 0.6 1.3 36.4

8.3 4.6 8.0 1.3 12.2 0.7 1.3 36.5

8.3 4.6 6.6 1.6 12.2 0.9 1.3 35.6

8.3 4.6 6.6 1.8 12.2 1.1 1.3 36.0

8.3 4.6 6.6 2.1 11.6 1.3 1.5 36.0

8.3 4.6 6.6 2.4 11.5 1.6 1.5 36.4

8.3 4.7 6.6 2.7 11.5 1.8 1.6 37.1

8.3 4.7 5.7 3.0 11.7 2.0 1.7 37.1

8.3 4.7 5.7 3.4 11.5 2.2 1.8 37.5

8.3 4.7 5.7 3.7 11.6 2.3 1.8 38.0

8.3 4.7 5.7 4.0 11.6 2.5 1.9 38.7

8.3 4.7 5.7 4.3 11.6 2.6 1.9 39.1

8.3 4.7 5.7 4.6 11.7 2.6 1.9 39.5

8.3 4.7 5.2 4.9 12.3 2.6 1.9 40.0

8.3 4.7 5.2 5.2 13.0 2.5 1.8 40.7

167


Appendix E

3A, SND GW Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

3B, SND GW Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Subtotal

2014

6.7 4.0 11.7 0.3 5.3 5.3 0.1 1.5 34.9

6.7 4.6 11.3 0.3 5.4 5.2 0.1 1.6 35.2

7.9 4.6 10.3 0.4 5.7 5.1 0.2 1.7 35.9

7.9 4.6 9.1 0.4 5.7 6.9 0.3 1.7 36.6

8.0 4.6 9.1 0.5 5.7 6.9 0.4 1.9 37.1

8.3 4.6 8.4 0.5 5.7 7.5 0.6 1.9 37.5

8.3 4.7 8.0 0.7 6.5 7.2 0.7 1.7 37.8

8.3 4.7 8.0 0.9 6.5 7.2 0.9 1.7 38.2

8.3 4.7 8.0 1.1 6.8 7.2 1.1 1.8 38.9

8.3 4.7 8.0 1.3 7.3 6.5 1.4 1.8 39.2

8.3 4.7 8.0 1.5 7.2 6.5 1.6 1.8 39.6

8.3 4.7 8.0 1.7 7.2 6.5 1.9 1.9 40.2

8.3 4.7 7.9 1.9 7.3 6.5 2.0 1.9 40.7

8.3 4.7 7.9 2.2 8.1 5.8 2.2 1.9 41.1

8.3 4.7 7.9 2.3 8.3 5.8 2.4 1.9 41.7

8.3 4.7 7.9 2.6 8.7 5.8 2.5 1.8 42.4

8.3 4.7 7.9 2.8 8.7 5.8 2.7 1.8 42.8

8.3 4.7 7.9 3.3 8.7 5.8 2.8 1.8 43.3

8.3 4.7 7.5 3.5 9.3 5.8 2.9 1.8 43.8

8.3 4.7 7.5 3.5 10.1 5.8 2.8 1.8 44.5

6.7 4.0 11.7 0.3 5.3 5.3 0.1 1.5 34.9

2015

6.7 4.6 11.3 0.3 5.4 5.2 0.1 1.6 35.2

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

3C, SND GW Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

3D, SND GW Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

3E, SND GW Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 5.3 5.3 0.1 1.5 34.9

6.7 4.0 11.7 0.3 5.3 5.3 0.1 1.5 34.9

6.7 4.0 11.7 0.3 5.3 4.6 0.1 1.5 34.2

6.7 4.6 11.3 0.3 5.4 5.2 0.1 1.6 35.2

6.7 4.6 11.3 0.3 5.4 5.2 0.1 1.6 35.2

6.7 4.6 11.3 0.3 5.4 4.5 0.1 1.5 34.5

7.9 4.6 10.3 0.4 5.7 5.1 0.2 1.7 35.9

7.9 4.6 10.3 0.4 5.7 5.1 0.2 1.7 35.9

7.9 4.6 10.3 0.4 5.7 5.1 0.2 1.7 35.9

7.9 4.6 10.3 0.4 5.7 4.6 0.2 1.6 35.2

7.9 4.6 9.1 0.4 5.7 6.8 0.3 1.8 36.6

7.9 4.6 9.1 0.4 5.7 6.8 0.3 1.8 36.6

7.9 4.6 9.1 0.4 5.7 6.8 0.3 1.8 36.6

7.9 4.6 9.1 0.4 5.7 6.3 0.3 1.6 35.9

8.0 4.6 9.1 0.5 5.7 6.9 0.4 1.9 37.1

8.0 4.6 9.1 0.5 5.7 6.8 0.4 1.9 37.1

8.0 4.6 9.1 0.5 5.7 6.8 0.4 1.9 37.1

8.0 4.6 9.1 0.8 5.7 6.1 0.4 1.6 36.4

8.3 4.6 8.4 0.5 5.7 7.5 0.6 1.9 37.5

8.3 4.6 8.4 0.7 5.7 7.2 0.6 1.9 37.4

8.3 4.6 8.4 0.5 5.7 7.5 0.6 1.9 37.5

8.3 4.6 8.4 1.2 5.7 6.5 0.6 1.5 36.8

8.3 4.7 8.0 0.7 6.1 7.2 0.7 1.9 37.6

8.3 4.7 8.0 0.9 5.9 7.2 0.7 1.9 37.6

8.3 4.7 8.0 0.5 6.5 7.2 0.7 1.9 37.8

8.3 4.6 8.0 1.7 5.7 6.5 0.7 1.4 36.9

8.3 4.7 8.0 0.9 6.5 7.2 0.9 1.9 38.4

8.3 4.7 8.0 1.1 6.1 7.2 0.9 1.9 38.1

8.3 4.7 8.0 0.7 6.5 7.2 0.9 1.9 38.2

8.3 4.7 8.0 1.9 5.7 6.5 0.9 1.5 37.5

8.3 4.7 8.0 1.1 6.6 7.2 1.1 1.9 38.9

8.3 4.7 8.0 1.3 7.2 6.4 1.1 1.9 38.9

8.3 4.7 8.0 0.9 6.8 7.2 1.1 1.9 38.9

8.3 4.7 8.0 2.3 5.8 6.5 1.1 1.6 38.2

8.3 4.7 8.0 1.3 7.2 6.5 1.4 1.9 39.3

8.3 4.7 8.0 1.5 7.0 5.7 1.4 1.9 38.5

8.3 4.7 8.0 1.1 7.4 6.5 1.4 1.9 39.2

8.3 4.7 8.0 2.6 6.4 5.8 1.3 1.7 38.8

168

8.3 4.7 8.0 1.5 7.2 6.5 1.6 1.9 39.7

8.3 4.7 8.0 1.7 6.9 5.7 1.7 1.9 38.9

8.3 4.7 8.0 1.3 7.2 6.5 1.7 1.9 39.6

8.3 4.7 8.0 2.9 6.6 5.8 1.6 1.8 39.6

8.3 4.7 8.0 1.8 7.2 6.5 1.9 1.9 40.2

8.3 4.7 8.0 2.0 7.0 5.7 1.9 1.9 39.4

8.3 4.7 8.0 1.6 7.2 6.5 2.1 1.9 40.2

8.3 4.7 8.0 3.2 6.5 5.8 1.8 1.9 40.2

8.3 4.7 7.9 2.0 7.3 6.5 2.0 1.9 40.7

8.3 4.7 7.9 2.2 7.1 5.7 2.1 1.9 39.9

8.3 4.7 7.0 1.8 8.1 6.5 2.4 1.9 40.7

8.3 4.7 7.0 3.6 7.4 5.8 2.0 1.9 40.7

8.3 4.7 7.9 2.2 8.1 5.8 2.2 1.9 41.1

8.3 4.7 7.9 2.4 7.2 5.7 2.3 1.9 40.4

8.3 4.7 7.0 2.0 8.0 6.5 2.8 1.9 41.2

8.3 4.7 7.0 3.9 7.4 5.8 2.2 1.9 41.2

8.3 4.7 7.9 2.4 8.3 5.8 2.4 1.9 41.7

8.3 4.8 7.9 2.6 7.3 5.7 2.4 1.9 40.9

8.3 4.7 7.0 2.2 8.7 5.8 3.1 1.9 41.7

8.3 4.7 7.0 4.2 7.6 5.8 2.3 1.8 41.7

8.3 4.7 7.9 2.6 8.7 5.8 2.5 1.8 42.4

8.3 4.8 7.9 2.8 8.3 5.1 2.6 1.9 41.6

8.3 4.7 7.0 2.4 8.9 5.8 3.4 1.9 42.4

8.3 4.7 7.0 4.5 7.9 5.8 2.5 1.6 42.4

8.3 4.7 7.9 2.8 8.7 5.8 2.7 1.8 42.8

8.3 4.8 7.9 3.1 8.3 5.1 2.8 1.9 42.1

8.3 4.7 7.0 2.5 8.7 5.8 3.8 1.9 42.9

8.3 4.7 7.0 4.8 7.9 5.8 2.7 1.6 42.9

8.3 4.7 7.9 3.2 8.7 5.8 2.8 1.8 43.3

8.3 4.8 7.9 3.5 8.3 5.1 2.9 1.9 42.6

8.3 4.7 7.0 2.8 8.7 5.8 4.2 1.9 43.5

8.3 4.7 7.0 5.1 7.9 5.8 2.7 1.6 43.3

8.3 4.7 7.5 3.5 9.3 5.8 2.9 1.8 43.8

8.3 4.8 7.5 3.6 9.0 5.1 2.9 1.9 43.0

8.3 4.7 6.6 3.3 8.7 5.8 4.4 1.9 43.8

8.3 4.7 6.6 5.5 8.4 5.8 2.8 1.7 43.8

8.3 4.7 7.5 3.6 9.8 5.8 2.8 1.8 44.4

8.3 4.8 7.5 3.5 10.5 5.1 2.9 1.9 44.4

8.3 4.7 6.6 3.3 9.3 5.8 4.6 1.9 44.6

8.3 4.7 6.6 5.7 8.8 5.8 2.7 1.8 44.5


Appendix E


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.4 0.1 1.3 34.6

6.7 4.6 11.3 0.3 10.1 0.1 1.3 34.5

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.4 11.4 0.3 1.3 35.0

8.0 4.6 9.1 0.5 11.2 0.4 1.3 35.2

8.3 4.6 8.4 0.5 12.2 0.6 1.3 35.9

8.3 4.7 8.0 1.0 12.2 0.7 1.3 36.2

8.3 4.7 6.6 1.2 12.2 0.9 1.3 35.2

8.3 4.7 6.6 1.4 12.2 1.1 1.3 35.6

8.3 4.7 6.6 1.6 11.5 1.4 1.3 35.3

8.3 4.7 6.6 1.8 11.5 1.7 1.3 35.8

8.3 4.7 6.6 2.0 11.5 1.9 1.3 36.3

8.3 4.7 5.4 2.2 11.6 2.1 1.5 35.8

8.3 4.7 5.4 2.4 11.5 2.3 1.6 36.1

8.3 4.7 5.4 2.6 11.4 2.4 1.7 36.5

8.3 4.7 5.4 2.8 11.5 2.6 1.8 37.1

8.3 4.7 5.4 3.0 11.4 2.8 1.8 37.4

8.3 4.7 5.4 3.2 11.5 2.9 1.8 37.8

8.3 4.7 5.0 3.3 12.2 2.9 1.8 38.2

8.3 4.7 5.0 3.7 12.3 2.9 1.8 38.7


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.4 0.1 1.3 34.6

6.7 4.6 11.3 0.3 10.1 0.1 1.3 34.5

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.4 11.4 0.3 1.3 35.0

8.0 4.6 9.1 0.5 11.2 0.4 1.3 35.2

8.3 4.6 8.4 0.5 12.2 0.6 1.3 35.9

8.3 4.7 8.0 0.9 12.2 0.7 1.3 36.1

8.3 4.7 6.6 1.1 12.2 0.9 1.3 35.2

8.3 4.7 6.6 1.3 12.2 1.1 1.3 35.6

8.3 4.7 6.6 1.5 11.5 1.4 1.3 35.3

8.3 4.7 6.6 1.8 11.5 1.7 1.3 35.8

8.3 4.7 6.6 2.0 11.5 1.9 1.3 36.3

8.3 4.7 5.4 2.2 11.6 2.1 1.5 35.8

8.3 4.7 5.4 2.4 11.5 2.3 1.6 36.1

8.3 4.7 5.4 2.6 11.4 2.4 1.7 36.5

8.3 4.7 5.4 2.8 11.5 2.6 1.8 37.1

8.3 4.7 5.4 3.0 11.4 2.8 1.8 37.4

8.3 4.7 5.4 3.2 11.5 2.9 1.8 37.8

8.3 4.7 5.0 3.3 12.2 2.9 1.7 38.2

8.3 4.7 5.0 3.7 12.5 2.9 1.6 38.7


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.4 0.1 1.3 34.6

6.7 4.6 11.3 0.3 10.1 0.1 1.3 34.5

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.4 11.4 0.3 1.3 35.0

8.0 4.6 9.1 0.5 11.2 0.4 1.3 35.2

8.3 4.6 8.4 0.5 12.2 0.6 1.3 35.9

8.3 4.7 8.0 1.1 12.2 0.7 1.3 36.3

8.3 4.7 6.6 1.3 12.2 0.9 1.3 35.4

8.3 4.7 6.6 1.5 11.5 1.1 1.3 35.0

8.3 4.7 6.6 1.7 10.7 1.4 1.3 34.7

8.3 4.7 6.6 2.0 10.7 1.7 1.3 35.3

8.3 4.7 6.6 2.2 10.7 2.0 1.5 35.9

8.3 4.8 5.7 2.4 10.9 2.2 1.6 35.8

8.3 4.8 5.7 2.6 10.7 2.4 1.7 36.1

8.3 4.8 5.7 2.8 10.7 2.5 1.8 36.6

8.3 4.8 5.7 3.0 10.8 2.7 1.9 37.1

8.3 4.8 5.7 3.2 10.7 2.9 1.9 37.4

8.3 4.8 5.7 3.4 10.8 3.0 1.9 37.8

8.3 4.8 5.2 3.5 11.4 3.0 1.9 38.2

8.3 4.8 5.2 3.7 11.8 3.0 1.9 38.7


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.3 0.1 1.5 34.6

6.7 4.2 11.3 0.3 10.3 0.1 1.6 34.5

7.9 4.2 10.3 0.4 10.2 0.2 1.7 34.9

7.9 4.2 9.1 0.4 11.3 0.3 1.8 35.0

8.0 4.3 9.1 0.5 11.2 0.4 1.9 35.4

8.3 4.2 8.4 0.5 12.2 0.6 1.9 36.1

8.3 4.3 8.0 1.0 12.2 0.7 1.9 36.4

8.3 4.3 5.8 1.2 12.2 0.9 1.9 34.6

8.3 4.3 5.8 1.4 12.2 1.1 1.9 35.0

8.3 4.3 5.8 1.6 11.5 1.4 1.9 34.7

8.3 4.3 5.8 1.8 11.5 1.7 1.9 35.3

8.3 4.4 5.8 2.0 11.5 2.1 1.9 35.9

8.3 4.4 4.6 2.2 12.0 2.4 1.9 35.8

8.3 4.4 4.6 2.4 11.8 2.8 1.9 36.1

8.3 4.4 4.6 2.6 11.7 3.1 1.9 36.6

8.3 4.4 4.6 2.8 11.7 3.4 1.9 37.2

8.3 4.4 4.6 3.0 11.4 3.8 1.9 37.4

8.3 4.4 4.6 3.2 11.4 4.2 1.9 37.9

8.3 4.4 4.2 3.3 11.7 4.4 1.9 38.2

8.3 4.4 4.2 3.3 12.2 4.6 1.9 38.9


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.4 0.1 1.3 34.6

6.7 4.6 11.3 0.3 10.1 0.1 1.3 34.5

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.5 11.3 0.3 1.3 35.0

8.0 4.6 9.1 0.6 11.2 0.4 1.3 35.3

8.3 4.6 8.4 0.8 12.2 0.6 1.3 36.2

8.3 4.7 8.0 1.0 12.2 0.7 1.3 36.2

8.3 4.7 5.8 1.3 12.2 0.9 1.3 34.6

8.3 4.7 5.8 1.6 12.2 1.1 1.3 35.0

8.3 4.7 5.8 1.9 11.5 1.4 1.3 34.9

8.3 4.7 5.8 2.2 11.5 1.6 1.3 35.4

8.3 4.7 5.8 2.5 11.5 1.9 1.5 36.1

8.3 4.7 4.6 2.8 11.7 2.1 1.6 35.8

8.3 4.7 4.6 3.1 11.5 2.3 1.7 36.1

8.3 4.7 4.6 3.4 11.4 2.4 1.7 36.6

8.3 4.7 4.6 3.7 11.4 2.5 1.9 37.1

8.3 4.7 4.6 4.0 11.4 2.7 1.7 37.5

8.3 4.7 4.6 4.3 11.4 2.8 1.7 37.9

8.3 4.7 4.2 4.6 11.8 2.9 1.7 38.2

8.3 4.7 4.2 4.9 12.2 2.8 1.8 38.8

169


Appendix E 5A, SND GW Nuclear Hydro Coal Renewables Gas EE DR Subtotal

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.0 0.1 1.3 34.2

6.7 4.2 11.3 0.3 10.1 0.1 1.3 34.2

7.9 4.2 10.3 0.4 10.2 0.2 1.3 34.6

7.9 4.2 9.1 0.4 11.2 0.3 1.3 34.5

8.0 4.3 9.1 0.5 11.2 0.4 1.3 34.8

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 6.6 0.5 12.2 0.9 1.3 34.1

8.3 4.3 6.6 0.5 12.2 1.1 1.3 34.3

8.3 4.3 6.6 0.5 11.9 1.3 1.5 34.3

8.3 4.3 6.6 0.5 11.5 1.6 1.6 34.4

8.3 4.4 6.6 0.5 11.5 1.8 1.6 34.6

8.3 4.4 5.4 0.6 12.4 2.0 1.7 34.8

8.3 4.4 5.4 0.8 12.2 2.2 1.7 35.0

8.3 4.4 5.4 1.0 12.2 2.3 1.7 35.3

8.3 4.4 5.4 1.2 12.2 2.5 1.7 35.7

8.3 4.4 5.4 1.4 12.2 2.6 1.6 35.9

8.3 4.4 5.4 1.6 12.2 2.6 1.6 36.1

8.3 4.4 5.0 1.7 12.6 2.6 1.7 36.3

8.3 4.4 5.0 1.8 13.0 2.5 1.7 36.6


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.0 0.1 1.3 34.2

6.7 4.2 11.3 0.3 10.1 0.1 1.3 34.2

7.9 4.2 10.3 0.4 10.2 0.2 1.3 34.6

7.9 4.2 9.1 0.4 11.2 0.3 1.3 34.5

8.0 4.3 9.1 0.5 11.2 0.4 1.3 34.8

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 6.6 0.5 12.2 0.9 1.3 34.1

8.3 4.3 6.6 0.5 12.2 1.1 1.3 34.3

8.3 4.3 6.6 0.5 11.9 1.3 1.5 34.3

8.3 4.3 6.6 0.5 11.5 1.6 1.6 34.4

8.3 4.4 6.6 0.5 11.5 1.8 1.6 34.6

8.3 4.4 5.4 0.6 12.4 2.0 1.7 34.8

8.3 4.4 5.4 0.8 12.2 2.2 1.7 35.0

8.3 4.4 5.4 1.0 12.2 2.3 1.7 35.3

8.3 4.4 5.4 1.2 12.2 2.5 1.7 35.7

8.3 4.4 5.4 1.4 12.2 2.6 1.6 35.9

8.3 4.4 5.4 1.6 12.2 2.6 1.6 36.1

8.3 4.4 5.0 1.7 12.6 2.6 1.7 36.3

8.3 4.4 5.0 1.8 13.0 2.5 1.7 36.6


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.0 0.1 1.3 34.2

6.7 4.2 11.3 0.3 10.1 0.1 1.3 34.2

7.9 4.2 10.3 0.4 10.2 0.2 1.3 34.6

7.9 4.2 9.1 0.4 11.2 0.3 1.3 34.5

8.0 4.3 9.1 0.5 11.2 0.4 1.3 34.8

8.3 4.2 8.4 0.5 12.2 0.6 1.3 35.5

8.3 4.3 8.0 0.5 12.2 0.7 1.3 35.3

8.3 4.3 8.0 0.5 12.2 0.9 1.3 35.5

8.3 4.3 8.0 0.5 11.5 1.1 1.3 34.9

8.3 4.3 8.0 0.5 10.7 1.3 1.3 34.4

8.3 4.3 8.0 0.5 10.7 1.6 1.3 34.7

8.3 4.3 8.0 0.5 10.7 1.8 1.4 34.9

8.3 4.4 7.0 0.7 10.9 2.0 1.5 34.8

8.3 4.4 7.0 1.0 10.7 2.2 1.5 35.0

8.3 4.4 7.0 1.1 10.6 2.3 1.5 35.3

8.3 4.4 7.0 1.3 10.6 2.5 1.5 35.7

8.3 4.4 7.0 1.5 10.6 2.7 1.5 36.0

8.3 4.4 7.0 1.7 10.6 2.7 1.5 36.2

8.3 4.4 6.6 1.8 10.9 2.7 1.6 36.3

8.3 4.4 6.6 1.9 11.1 2.6 1.7 36.6


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 9.9 0.1 1.5 34.2

6.7 4.2 11.3 0.3 9.9 0.1 1.6 34.2

7.9 4.2 10.3 0.4 10.2 0.2 1.7 34.9

7.9 4.2 9.1 0.4 11.2 0.3 1.8 35.0

8.0 4.3 9.1 0.5 11.2 0.4 1.9 35.4

8.3 4.2 8.4 0.5 12.2 0.6 1.9 36.1

8.3 4.3 8.0 0.5 12.2 0.7 1.9 35.9

8.3 4.3 6.6 0.5 12.2 0.9 1.9 34.7

8.3 4.3 6.6 0.5 12.2 1.1 1.9 34.9

8.3 4.3 6.6 0.5 11.5 1.4 1.9 34.4

8.3 4.3 6.6 0.5 11.5 1.7 1.9 34.8

8.3 4.3 6.6 0.5 11.5 2.1 1.9 35.1

8.3 4.4 5.4 0.5 11.9 2.4 1.9 34.8

8.3 4.4 5.4 0.5 11.8 2.8 1.9 35.1

8.3 4.4 5.4 0.5 11.8 3.1 1.9 35.4

8.3 4.4 5.4 0.6 11.7 3.4 1.9 35.7

8.3 4.4 5.4 0.6 11.4 3.8 1.9 35.8

8.3 4.4 5.4 0.8 11.4 4.2 1.9 36.3

8.3 4.4 5.0 0.9 11.4 4.4 1.9 36.3

8.3 4.4 5.0 1.0 11.6 4.6 1.9 36.8


2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

6.7 4.0 11.7 0.3 10.0 0.1 1.3 34.2

6.7 4.6 11.3 0.3 9.9 0.1 1.3 34.3

7.9 4.6 10.3 0.4 10.2 0.2 1.3 34.9

7.9 4.6 9.1 0.4 11.2 0.3 1.3 34.9

8.0 4.6 9.1 0.6 11.2 0.4 1.3 35.3

8.3 4.6 8.4 0.7 12.2 0.6 1.3 36.1

8.3 4.6 7.1 0.8 12.2 0.7 1.3 35.1

8.3 4.6 5.8 1.0 12.2 0.9 1.3 34.1

8.3 4.6 5.8 1.3 12.2 1.1 1.3 34.5

8.3 4.6 5.8 1.6 11.5 1.3 1.3 34.3

8.3 4.6 5.8 1.9 11.5 1.5 1.3 34.9

8.3 4.7 5.8 2.2 11.5 1.7 1.3 35.4

8.3 4.7 4.6 2.5 11.7 1.8 1.3 34.8

8.3 4.7 4.6 2.8 11.5 2.0 1.3 35.1

8.3 4.7 4.6 3.0 11.4 2.1 1.3 35.3

8.3 4.7 4.6 3.3 11.4 2.1 1.3 35.6

8.3 4.7 4.6 3.6 11.4 2.1 1.3 36.0

8.3 4.7 4.6 3.9 11.4 2.1 1.4 36.3

8.3 4.7 4.1 4.1 11.5 2.0 1.5 36.2

8.3 4.7 4.1 4.4 11.6 1.9 1.6 36.6

170

Appendix F

Method for Computing Environmental Metrics

171


Appendix F

Process

Method

In developing the criteria for the environmental impact metrics, TVA wanted to create a set of metrics representative of the trade-offs between energy resources rather than identifying a single resource with “best” environmental performance. By considering air, water and waste in the IRP scorecard, coupled with the broader qualitative discussion of anticipated environmental impacts in the EIS, a robust comparison of the environmental footprint of the planning strategies better informed the selection of the recommended strategy.

The environmental impact metrics can be grouped into two broad categories: Scoring metrics – these metrics will be used in the strategy scorecard to assess the performance of a given set of portfolios created by modeling that strategy across the scenarios used in the study. Reporting metrics – will be computed and included in the IRP report as informational or supplemental measures to help clarify or expand on the insights. Three environmental impact metrics for air, water and waste were selected for scoring and two, air and waste, for reporting metrics. The scoring metrics are shown in Figure 1.

Scoring Metric

Definition

CO2 Avg Tons

The annual average tons of CO2 emitted over the study period

Water Consumption

The annual average gallons of water consumed over the study period The annual average quantity of coal ash, sludge & slag projected based on energy production in each portfolio

Waste

Figure F‑1: Scoring Metrics

Category


Scoring Metric

Formula

CO2 (MMTons)

=

Average Annual Tons of CO2 Emitted During Planning Period

Water Consumption (Million Gallons)

=

Average Annual Gallons of Water Consumed During Planning Period

Waste (MMTons)

=

Average Annual Tons of Coal Ash and Scrubber Residue During Planning Period

Figure F‑2: Scoring Metric Formulas 172


Appendix F

The two reporting metrics are shown in Figure 3. Reporting Metric

Definition

CO Intensity

The CO emissions expressed as an emission intensity; computed by 2 dividing emissions by energy generated

Spent Nuclear Fuel Index

A measure of the quantity of spent nuclear fuel that is projected to be generated based on energy production in each portfolio

2

Figure F‑3: Reporting Metrics The formulas for the reporting metrics are shown in Figure 4.

Category

Reporting Metric

Formula


=

Spent Nuclear Fuel Index (Tons)

=


Figure F‑4: Reporting Metric Formulas

173

Tons CO2 (2014-2033) GWh Generated (2014-2033) Expected Spent Fuel Generated During Planning Period


Appendix F

Strategy Performance: Air Impact Metric

80,000 70,000

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

14

5A

20,000

4A

30,000

3A

40,000

2A

1A

50,000

60,000

20

Annual CO2 Emissions in Thousand Tons

CO2 Scoring metric results:

Year

80,000 70,000 60,000

1B

Strategy B-Meet an Emission Target

174

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

14

Year

5B

20,000

4B

30,000

3B

40,000

2B

50,000

20


Strategy A-The Reference Plan



Appendix F

80,000 70,000 60,000

1C

50,000

2C

40,000

3C

30,000

4C

5C

20

32

20

30

28 20

26 20

24 20

22 20

20

20

18 20

16 20

20

14

20,000

Year

80,000 70,000 60,000

1D

50,000

2D

40,000

3D

30,000

4D 5D

Year

Strategy D-Maximize Energy Efficiency

175

2 20 3

0 20 3

8 20 2

6 20 2

4 20 2

2 20 2

0 20 2

8 20 1

20 1

20 1

6

20,000

4


Strategy C-Focus on Long-term, Market Resources


80,000 70,000 60,000

1E

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

20

5E

20,000

4E

30,000

3E

40,000

2E

50,000

14


Appendix F

Year

Strategy E-Maximize Renewables

1,000 900 800 700 600 500 400 300

1A 2A 3A

Year

Strategy A – The Reference Plan

176

2 20 3

0 20 3

28 20

6 20 2

4 20 2

2 20 2

0 20 2

18 20

6 20 1

4

4A

20 1

Annual CO2 Emissions Rate in lbs/mwh

CO2 Reporting metric results:

5A


1,000 900 800 700 600 500 400 300

1B 2B 3B

20 32

20 30

20 28

20 26

20 24

20 22

20 20

20 18

20 16

4B

20 14


Appendix F

5B

Year

1,000 900 800 700 600 500 400 300

1C 2C 3C

Year


177

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

14

4C

20



5C


1,000 900 800 700 600 500 400 300

1D 2D 3D

32 20

30 20

28 20

26 20

24 20

22 20

20

20

18 20

16 20

14

4D

20


Appendix F

5D

Year

1,000 900 800 700 600 500 400 300

1E 2E 3E

Year


Air Impact Metric Observations: • CO2 emissions vary largely by scenario but decline over time for all strategies • Strategies A,B and C have similar CO2 emission profiles across the scenarios, coming in about 3 percent above Strategy D and about 10 percent above Strategy E • Strategy E achieves the lowest intensity at 296 tons/ GWh, which is about 10 percent lower than A, B and C and about 8 percent lower than D

178

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

14

4E

20



5E


Appendix F

Strategy Performance: Water Impact Metric

75,000 65,000 55,000 45,000

1A 2A 32 20

20

30

28 20

26 20

24 20

22 20

20

20

20

18

16 20

14

3A

20

Annual Water Consump0on in Million Gallons

Scoring metric results:

Year

4A 5A

70,000 65,000

1B

60,000

2B

55,000

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

14

4B

3B

45,000

50,000

20



5B

Year

70,000 65,000 60,000 55,000 50,000 45,000

1C 2C

Year


179

20 32

20 30

20 28

20 26

20 24

20 22

20 20

20 18

3C 20 16

20 14



4C 5C


70,000 65,000

1D

60,000

20 32

20 30

20 28

20 26

4D

20 24

45,000 20 22

3D 20 20

50,000 20 18

2D

20 16

55,000

20 14


Appendix F

5D

Year

70,000 65,000 60,000 55,000 50,000 45,000

1E 2E

Year


Water Impact Metric Observations: • Average water consumption declines over time in all strategies. • Variation across scenarios within a particular strategy ranges from 10.5 percent for Strategies A/B to 13.8 percent for Strategy D. This is largely driven by the variation in load growth in the different scenarios. • Average water consumption across the five strategies ranges from 56,960 for Strategy E to 59,210 for Strategy C or 2,250 million gallons. This represents a variation of about 4 percent.

180

4E

32 20

30 20

28 20

26 20

20

24

22 20

20 20

18 20

16 20

14

3E

20



5E


Appendix F

Strategy Performance: Waste Impact Metric

Annual Produc4on in Tons

Scoring metric results: 6000000 5000000 1A

4000000

2A

3000000

3A

2000000

4A

1000000

5A

32

20

20

30

20

28

26

20

24

20

22

20

20

20

18 20

16 20

20

14

0



6000000 5000000 1B

4000000

2B

3000000

3B

2000000

4B

1000000

5B


181

32 20

30 20

28 20

26 20

24 20

20

22

20 20

18 20

16 20

20

14

0



Appendix F

6000000 5000000

1C

4000000

2C

3000000

3C

2000000

4C

1000000

5C

32 20

30 20

28 20

26 20

24 20

22 20

20 20

18 20

16 20

20

14

0



6000000 5000000

1D

4000000

2D

3000000

3D

2000000

4D

1000000

5D


182

20 32

20 30

28 20

20 26

20 24

20 22

20 20

18 20

20 16

20 14

0



Appendix F

6000000 5000000

1E

4000000

2E

3000000

3E

2000000

4E

1000000

5E


Waste Impact Metric Observations: The trends in the production of high-level waste, which is primarily spent nuclear fuel and other fuel assembly components, parallel those of nuclear fuel requirements and are the same for all alternative strategies and average 149.05 Tons/Year.

183

32 20

30 20

20

28

26 20

24 20

22 20

20 20

18 20

16 20

20

14

0

Appendix G

Method for Computing Valley Economic Impacts

185


Appendix G

Background

that dampen it. REMI is a general equilibrium model used by TVA for well over a decade and is currently in use by over 100 universities, state and local governments, utilities, and consulting firms throughout the U.S. and Europe. TVA’s model has been tailored to the TVA Region by county and optimized to capture the inter-industry and inter-regional linkages with surrounding counties and the rest of the United States. As shown in Figure G-1, the “direct effects,” i.e. changes in TVA expenditures and retail electricity bills, are input into REMI, which capture any multiplier effects and interactions within the regional economy.

Since the TVA Act promotes agricultural and industrial development as a core TVA responsibility, the economic well-being of Tennessee Valley (hereafter Valley) residents has been part of the TVA’s mission since 1933. In keeping with TVA’s core mission, the IRP scoring process incorporates a single economic impact metric for each strategy of every scenario under consideration. Per capita income is calculated in order to assess the relative impact of each strategy on the general economic conditions of in the TVA Region. This metric is used as one input into the overall IRP Scorecard used to evaluate alternative strategies. As second metric, Valley employment is also included in this appendix but is not part of the scorecard.

Process Overview The U.S. Bureau of Economic Analysis provides a broad measure of per capita income that reflects not just wage income but total compensation, such as employer contributions to health insurance and retirement accounts. Additionally, it includes other income sources, such as dividends and transfer payments. Thus, per capita income provides a single metric that broadly reflects the general economic well-being of Valley residents and is readily understandable and relatable. It is also one that will reflect the net effect of each strategy’s change in expenditures and electricity bills. Increases in TVA expenditures on labor, equipment and construction materials stimulate the economy. At the same time, increases in consumers’ electricity bills required to fund those operations and construction activities, reduce consumers’ disposable income. Lower disposable income limits consumer purchases on goods and services in the TVA Region. Since strategies that involve increasing in-Valley expenditures tend to require higher electricity bills, their impacts tend to be offsetting.

Figure G‑1: Input and Output Impacts Strategy A of each scenario serves as the Reference Plan, so each strategy within each scenario is compared to Strategy A. Thus, increases in expenditures are only entered into REMI to the extent

The PI+ Model by Regional Economic Models, Inc., hereafter referred to as REMI, is used to model the multiplier effects of each strategy’s expenditures that stimulate the regional economy and its electrical bills

186


Appendix G

that they exceed Strategy A’s expenses. In this way REMI’s outputs are the impact on per capita income relative to the Reference Plan of each scenario.

While there are ongoing national codes and standards that increase energy efficiency, TVA implements programs that expedite the adoption of energy efficient appliances and insulation that are over and above the minimum required. The economic impact of TVA investments in energy efficiency programs are modeled as eight new jobs in the TVA Region for each $1 million spent. Of the jobs created 20 percent fall in the utility industry, 20 percent in construction industry, and 60 percent in professional/scientific employment categories. All differences from the Reference Plan are annual values, so changes in per capita income are generated by year. The per capita income output models the trajectory of economic impacts over time. In order to rank and compare alternative strategies, the present value of the changes in per capita income is evaluated with a 2 percent discount rate from 2014 to 2033. A low 2 percent discount rate is employed, because the changes in per capita income were previously adjusted for inflation. Selecting a rate as high as 8 percent does not, however, materially impact the strategy rankings. The results are presented below for non-farm employment as well.

Methodology Each strategy has a different annual revenue requirement needed to fund its construction, generation, and energy efficiency programs. The difference between the Reference Plan and the other strategies’ revenue requirements are modeled as changes in the electricity bill for residential, commercial, and industrial customers. Ultimately, rate payers must fund any increase in TVA expenditures. While increases in a strategy’s revenue requirements tend to reduce consumers’ ability to purchase goods and services, an increase in TVA expenditures stimulates economic activity, at least to the extent that they are purchased within the TVA Region. Expenditures that are almost exclusively sourced outside the TVA Region, such as fuel or purchased wind power from the Midwest, are excluded from TVA Region expenditures.

Overall Findings

Since not all types of expenses have identical economic impacts, REMI was used to separately model the impact of renewable construction, non-renewable construction, non-fuel operation and maintenance (O&M), and energy efficiency expenses. In this way REMI identifies the ability of the TVA Region’s economy to supply the necessary inputs and to what extent they must be sourced outside the region. Since most new construction expenses are likely to be natural gas-fired power plants, REMI’s custom construction industry for natural gas-fired power plants was incorporated into the analysis. Similarly, since most new renewable construction in the TVA Region will be solar installations, REMI’s custom industry for solar plant construction was used. This delineation between types of construction expenditures enhanced the accuracy of the results and followed directly from stakeholder feedback after completion of the 2011 Integrated Resource Plan.

Figure G-2 provides changes in the TVA Region’s per capita income caused by each strategy. The difference in all scenarios for all strategies is quite small. From 2014 to 2033 the average percentage change in per capita income ranged from -0.01% to 0.03%. The results are expected to be small for several reasons. First, TVA’s revenue is a small percentage of the total TVA Region economy. In 2015, TVA’s revenues are expected to approach $11 billion, but the entire TVA Region economy is almost $430 billion. Second, all the proposed strategies are similar approaches to supplying the region’s power needs. Changing from one approach to another should not result in significant impacts on the economy as a whole.

187


Appendix G

Per Capita Income* 2014-2033

Scenario 1

Scenario 2

Scenario 3

Scenario 4

Scenario 5

Avg. of Annual % Changes

Current

Stagnant

Growth

De-Carbonized

Distributed

from Reference Plan

Outlook

Economy

Economy

Future

Marketplace

B - Meet Emission Target

0.00%

0.01%

-0.01%

0.00%

0.00%

C - Focus on LT Market*

0.00%

0.01%

0.03%

0.01%

0.00%

D - Maximize EE

0.02%

0.02%

0.02%

0.02%

0.02%

E - Maximize Renewables

-0.01%

0.00%

0.00%

0.00%

-0.01%

Present Value of Per Capita

Income (2013$)

A- Reference Plan

$38,074

$36,206

$39,590

$37,502

$38,074


$38,074

$36,208

$39,588

$37,501

$38,074

C - Focus on LT Market* *

$38,073

$36,209

$39,602

$37,505

$38,073

D - Maximize EE

$38,080

$36,213

$39,597

$37,510

$38,081


$38,069

$36,204

$39,588

$37,502

$38,069

* U.S. Bureau of Economic Analysis definition reflects total compensation that

includes wages and benefits and transfer payments, such as Medicare and Medicaid. ** Full Name: Focus on Long-Term, Market-Supplied Resources

Figure G‑2: Results

Across the five scenarios, there are meaningfully different assumptions about economic conditions nationwide that impact the TVA Region’s standard of living. Per capita incomes are not, however, comparable across scenarios because the varying scenario assumptions generally overwhelm strategy-driven impacts.

of the changes in per capita income is $116 over and above what would have been available in the Reference Plan. To get a sense of what is driving the results, the changes in in-Valley expenditures are graphed alongside the changes in revenue requirements. Increasing in-Valley expenditures provide an economic stimulus, while increases in the revenue requirements dampen economic growth.

Detailed Results – Current Outlook Scenario

An important interpretation caveat is that REMI models different types of expenditures differently. Dollars spent on solar construction have a different impact from dollars spent on gas plant construction or energy efficiency. Nonetheless, comparing aggregate changes in in-Valley expenditures and revenue requirements can provide insights into the model’s result.

Scenario 1 -Current Outlook reflects the expected case assumptions about the general state of the economy and power markets. In this scenario we find that Strategy D, Maximize Energy Efficiency (EE), is the most beneficial. From 2014 to 2033 the present value

188


Appendix G

Figure G‑3: Current Outlook

189


Appendix G

Strategy 1B - In terms of either in-Valley expenditures or revenue requirements, Strategy1B is little changed from the Reference Plan.

expensive and some revenues are spent on out-ofValley wind generation. Compared to the Reference Plan, revenue requirements increase more than in-Valley expenditures.

Strategy 1C - Strategy 1C experiences slightly higher revenue requirements than expenditures, which depresses per capita income.

Detailed Results – Stagnant Economy Scenario

Strategy 1D - Strategy 1D involves greater in-Valley expenditures, especially after 2024. Although relatively expensive, the in-Valley EE expenditures do lower revenue requirements relative to expenditures.

The Stagnant Economy scenario models a world in which economic growth fails to materialize as expected. Most strategies have a marginally positive impact, but the benefit of in-Valley EE expenditures gives Strategy D the largest gain.

Strategy 1E - Renewable generation is relatively

190


Appendix G

Figure G‑4: Stagnant Economy

191


Appendix G

Strategy 2B - In-Valley expenditures increase, but revenue requirements are only marginally higher.

Detailed Results – Growth Economy Scenario

Strategy 2C - Increases in in-Valley expenditures are consistently greater than the increases in revenue requirements.

The Growth Economy Scenario models an environment of higher than expected economic growth and growing demand for power. This is the one scenario in which Strategy C, Focus on Long-Term, Market-Supplied Resources, provides the greatest benefit. This is largely driven by TVA’s presumed ability to secure 10-year PPA’s prior to the profitability of in-Valley solar generation. Strategy D that emphasizes energy efficiency programs is, however, the still second most beneficial.

Strategy 2D - Significant EE investments, modeled as all in-Valley jobs, result in higher in-Valley expenditures. Even though revenue requirements increase, in-Valley expenditures more quickly increase. Strategy 2E - Higher cost renewables are less cost effective with limited sales growth and lighter carbon regulation.

192


Appendix G

Figure G‑5: Growth Economy

193


Appendix G

Strategy 3B - In-Valley expenditures are generally lower and revenue requirements are flat. Strategy 3C - Revenue requirements are driven down because TVA is able to sign 10-year Purchased Power Agreements with gas-fired power plant operators but afterwards builds solar generation as technological improvements make solar power more efficient. Strategy 3D - EE expenditures, modeled as in-Valley jobs, ramp up dramatically beginning in 2023 and create a significant stimulus effect that more than offsets the increased cost. Strategy 3E - Since wind is primarily located outside the Valley, focusing on renewables involves Valley

residents paying for out-of-Valley generation. Through 2024 revenue requirements are generally greater than in-Valley expenditures.

Detailed Results – De-Carbonized Future Scenario The De-Carbonized Future Scenario models a regulatory environment in which there are significant carbon taxes that impact the relative efficiency of alternative strategies. As in all but one scenario, the stimulus impact of Strategy D’s in-Valley EE investments generates a marginally more positive economic impact than the other strategies.

194


Appendix G

Figure G‑6: De-Carbonized Future

195


Appendix G

Strategy 4B - Strategy B’s revenue requirements and expenditures are little changed.

wind expenditures peak in 2019. After 2020 increases in Non-Fuel O&M and in-Valley Solar Construction expenses provide a stimulus.

Strategy 4C - Revenue requirement growth is moderated by limited growth in the cost of purchased power. Expenses increase with in-Valley solar construction after 2019 and significant EE in 2024 & 2025.

Detailed Results – Distributed Marketplace Scenario The Distributed Marketplace Scenario models a world in which the economic and technological changes facilitate a shift toward distributed power generation. In this scenario, Strategy D offers the only approach that improves upon the Reference Plan.

Strategy 4D - Significant in-Valley EE expenditures begin in 2024. Strategy 4E - Revenue requirements in 2016 through 2022 exceed in-Valley expenditures as out-of-Valley

196


Appendix G

Figure G‑7: Distributed Marketplace

197


Appendix G

Other Reportable Metric

Strategy 5B - In-Valley expenditures and revenue requirements are little changed.

Although not used in the analysis directly, percentage changes in Nonfarm employment from the Reference Plan are presented in this Figure G-8. Like changes in per capita income, changes in nonfarm employment are very small.

Strategy 5C – Changes in revenue requirements generally exceed in-Valley expenditures. Strategy 5D - EE investments lift in-Valley expenditures above increases in revenue requirements. Strategy 5E - Focusing on expensive --and often out-of-the-Valley wind-- renewables increases revenue requirements faster than in-Valley expenditures.

Conclusion There are multiple approaches to meeting the TVA region’s power needs. This analysis compared the economic impact of alternative strategies to that of the Reference Plan for every scenario. Each strategy involved changing the level of in-Valley expenditures and the magnitude of electricity bills required to satisfy each strategy’s funding needs. Using REMI’s PI+ general equilibrium model tailored to the TVA service territory, the impact on per capita income of alternative strategies for meeting power demand was evaluated. By using custom industry models and base REMI capabilities, the impacts of different types of expenditures (e.g., renewable construction, non-renewable construction, non-fuel O&M, etc.) were modeled explicitly. Under most scenarios Strategy D, Maximize EE, generated the largest gains in per capita income over and above the Reference Plan. EE expenditures disproportionately remain in the Valley and dampen future electricity costs. Both factors tend to improve the relative performance of Strategy D. That being said, the impact of all alternative strategies on per capita income was exceptionally small. Across all scenarios and strategies the average percentage change in per capita income from 2014 through 2033 ranged from -0.01% to 0.03%. The present value of the stream of annual differences is small as well. Over a 20-year period, the “Maximize EE” strategy provides an additional benefit whose present value ranges from $116 to $158.

198


Appendix G

NonFarm Employment 2014-2033

Scenario 1

Scenario 2

Scenario 3

Scenario 4

Scenario 5

Avg. of Annual % Changes

Current

Stagnant

Growth

De-Carbonized

Distributed

from Reference Plan

Outlook

Economy

Economy

Future

Marketplace


0.00%

0.03%

-0.01%

0.00%

0.00%


0.00%

0.04%

0.05%

0.02%

0.00%

D - Maximize EE

0.06%

0.11%

0.06%

0.07%

0.08%


-0.02%

0.02%

-0.01%

0.00%

-0.02%

Annual Average

(Thousands)

A- Reference Plan

4,338

3,837

4,717

4,189

4,338


4,338

3,839

4,717

4,188

4,338


4,338

3,839

4,720

4,189

4,338

D - Maximize EE

4,341

3,841

4,720

4,192

4,342


4,337

3,838

4,717

4,188

4,337

* Full Name: Focus on Long-Term, Market-Supplied Resources

Figure G‑8: Nonfarm Employment

199

tva.com/irp